Title: Requirements for a model of customer environmental behavior and a theory of customer environmental perception
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098384/00001
 Material Information
Title: Requirements for a model of customer environmental behavior and a theory of customer environmental perception
Physical Description: xii, 190 leaves. : illus. ; 28 cm.
Language: English
Creator: Grossbart, Sanford Lyle, 1943-
Publication Date: 1972
Copyright Date: 1972
Subject: Consumers -- Mathematical models   ( lcsh )
Marketing thesis Ph. D   ( lcsh )
Dissertations, Academic -- Marketing -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Thesis: Thesis -- University of Florida.
Bibliography: Bibliography: leaves 169-189.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098384
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000585154
oclc - 14192098
notis - ADB3786


This item has the following downloads:

PDF ( 7 MBs ) ( PDF )

Full Text




A Dissertation Presented To The Graduate Council Of
The University of Florida
In Partial Fulfillment Of The Requirements For The
Degree of Doctor of Philosophy



L O 6 O

Sa qqTa,doo


This study could not have been completed without the

aid of many individuals. Sincere appreciation is offered

to the members of the supervisory committee, Dr. R. B.

Thompson, Dr. A. A. Anderson, Dr. J. B. McFerrin, Dr. I. J.

Goffman, and Dr. R. H. Blodgett. Dr. R. B. Settle is

thanked for his criticism and guidance concerning method-

ology and statistical analysis.

In addition, the author acknowledges a sense of

special gratitude to Dr. Thompson, for introducing him to

the meaning and social significance of marketing theory,

and Dr. Anderson, for introducing him to the meaning of

marketing knowledge. Words can never repay these debts.

Finally, Harriet Grossbart deserves thanks for her

research and administrative aid, for her understanding

and courage, and, most of all, for her tolerance.




ACKNOWLEDGMENTS . . . . . . . iii

LIST OF TABLES . . . . . . . . . vi

LIST OF FIGURES . . . . . . . . . vii

ABSTRACT . . . . . . . . .. . ix


I INTRODUCTION . . . . . . . 1

Urban Planning and Environmental
Malfunction . . . . . . . 4
Behavioral Science Contributions . .. 7
The Need for a Model of Customer Spa-
tial Behavior . . . . . 10
Customer Spatial Behavior, Environmental
Behavior, and the Requirements for a
Model of Environmental Behavior . .. 10

Ego-Centered Environmental Behavior . 12
Reciprocity and the Customer-
Environment Interchange . . . 19
Dual Definition of the Environmental
Situation ..... . . 30
Contextual and Stimulus Properties
of the Environment . 41
A Four-Element Intersystem Congruence
Model . . .. . . . .. 58

PERCEPTION . . . . . . . 66

Purpose . . . . .. 68
Related Research ...... 69
An Exploratory Model for the Measurement
of Customer Environmental Maps . . 73

Basic Assumptions . . . . .. 74
Ordering Mechanisms . . . . .. 75
The Nature of Customer Environmental
Perception . . . . .. 77
The Functions of a CEM . . . . 79



The CEM as a Theoretical Construct and
Experimental Behavioral Output Mode
Measurable Conceptual Metrics of the
CEM . . . . . . . .

Absolute location . . . ....
Relative location . . . . .
Extensity . . . . . . .
Barriers . . . . . . .
The cognitive structure of ESS
elements . . . . . .

Summary Statement and Definition of a CEM


S. .. 97


II METHODOLOGY . . . . . . .

The Questionnaire . . . . .

The Expected Pattern of Behavior
Salient Store Attributes
Store Ratings . . . . . .
Relative Location . . . . .
Absolute Location . . . . .
Demographic Profile . . . .

Respondents . . . . . . .
Field Procedure . . . . . .

IV RESULTS . . . . . . . .

Presentation and Analysis of Results

. 99

. 99

. 99
S. 105
. 110
. 110
. 119

. 119
. 122

S. 124

Demographic Data Profile . . . .
Cognitive Structure . . . . .
MGS Measures and Tests of Hypothesized
Relationships . . . . . .

Conclusions and Implications . . ..

The Evoked Store Set . . . ..
Attribute Salience and Rating . .
Mental Geographic Sets . . . ..


BIBLIOGRAPHY . . . . . . .










. . . a . .

* .


Table Page



STORE SET SIZE . . . . . . . 128



DIMENSIONS . . . . . . . . 134


VECTOR DIMENSIONS . . . . . . . 137




Figure Page

STORE SET . . . . . . . 101



MATRIX . . . . . . . 107

STORE RATING MATRIX . . . . . . 109

GEOGRAPHIC SET . . . . . . .. 111

GEOGRAPHIC SET . . . . . . .. 112

DISTANCE . . . . . . . 114


VENIENCE . . . . . . . 116



Figure Page


OF SYMMETRY . . . .. . ... 118



Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in
Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy



Sanford Lyle Grossbart

August, 1972

Chairman: Dr. R. B. Thompson
Major Department: Marketing

The purpose of this exploratory study was to deter-

mine the requirements for a model of customer environmental

behavior, develop a theory of customer environmental per-

ception, measure the correspondence between perceptions of

retail space and physical reality, and identify potential

sources of perceptual distortion.

Urban planning and the behavioral sciences provided

a framework for the identification of five requirements

for an environmental behavior model. The requirements

suggested that the description and analysis of environ-

mental behavior be ego-centered, that the relationship

between an individual and an environmental situation be

viewed as a reciprocity, involving a dual interchange,

that the environmental situation be viewed as possessing

objective and phenomenological features, that the environ-

mental situation must be seen as influencing environmental

behavior by acting as a behavioral context and/or as a

proximal and distal stimulus, and that the model relate

its basic elements, the individual, the environmental

situation, the environmental decision process, and envir-

onmental response behavior in terms of intersystem con-


A theory of customer environmental perception was

developed in terms of a customer environmental map con-

struct which was employed as an exploratory behavioral

output mode capable of eliciting customers' perceptions

of urban retail reality in terms of an array of points

representing stores. The points represented cognitively

structured information categories related to each other

in terms of a subjective set of geographic relations.

The study examined the relationship between cognitive

structure and the geographic ordering of stores. A stand-

ardized personal interview was administered to thirty-one

women. The specification of an expected pattern of behav-

ior regarding the nature of a forthcoming trip, a product

class, a purchase intention, and the extent of prior

knowledge made it possible to elicit a store set response.

Two types of measurements were made, those defining the

cognitive structure of stores, in terms of a vector indi-

cating the relative salience of store attributes and a

matrix indicating the relative ratings of stores in terms

of salient attributes, and those defining the subjective

geographic relationship between stores. The vector-matrix

products produced by each interview indicated store pref-

erence. Multiple relative and absolute location responses

were compared to objective location characteristics to

determine the magnitude and direction of perceptual dis-


Size variation and concentration of store sets, pos-

sibly related to sign-learning tendencies, variations in

environmental decision process stages, and the functional

role of stores, appeared to introduce some unexplained

complexities into the relationship between the cognitive

structure measure of preference and the geographic meas-

ures. Analysis of the results revealed that the magnitude

of relative location distortion was greater in the case

of less preferred stores and the direction of relative

location distortion was different for most and less pre-

ferred stores. There was partial support for the greater

magnitude of physical distance absolute location distor-

tion in the case of less preferred elements, the magnifi-

cation of the direction of time absolute location distor-

tion for closer and more distant most preferred stores,

the higher rated convenience of closer most preferred

stores, and the violation of the metric property of sym-

metry for physical distance absolute location perceptions.

The interpretation of results indicated that there

was considerable value in the future application and

continued refinement of the customer environmental map as

a model of customer environmental perception. It appeared

that such efforts would improve our understanding of the

environmental decision process and provide a valuable

input for marketing strategy and urban planning.



The role of the city and the nature of urban problems1

(Cox, 1962; Lazer, 1971, p. 440) have not been major concerns

in marketing thought. These factors have been grouped with

other environmental variables and conceptualized as a set

of uncontrollable forces to which the individual firm must

adapt. The city, however, has long since exceeded some

critical mass and become an environment rather than a tempo-

rary retreat from natural surroundings (Lynch and Rodwin,

1960; Wurster, 19631 Parr, 1966, p. 39). This has trans-

formed certain marketing variables, including the urban

retail structure, into environmental components. The impli-

cations of this transformation are especially significant

in an era of urban planning.

The distribution of urban retail facilities is a

product of entrepreneurial vision, custom, zoning, and the

like. The characterization of the city as a marketing

institution (Cox, 1962; Lazer, 1971, pp. 440-43) has led

1The urban (or city) area is not limited to elements
within conventional metropolitan political boundaries but
also includes suburban features, which represent the decen-
tralization of production and/or consumption activities
(see Schnore, 1956, and Cox, 1962).

to the development of hypotheses concerning retail struc-

ture which have included changes in the relative costs of

moving goods and people, consumer uncertainty, various

aspects of customer behavior, shifting retail location,

and changes in retail function as key variables (Proudfoot,

1937; Gist and Halbert, 1956, pp. 90-106; Berry and Garri-

son, 1958a; Huff, 1960; Boal, 1963; Mertes, 1964; Wagner,

1964; D. L. Thompson, 1966, 1967; Cohen and Lewis, 1967;

Van Handel, 1968). The absence of unified theory relating

urban retail structure (Durden and Marble, 1961; Eisenpress,

1965, p. 244) has not, however, retarded efforts to plan

and construct the urban complex and its retail component.

Societal pressure as well as the resource, structural, and

productivity dimensions of urban space have left no other

alternative (Wingo, 1963, p. 7; B. J. L. Berry, 1965; Fagin,

1967; Goodman and Freund, 1968; Morris, 1969a; Lansing,

Marans, and Zehner, 1970).

The paucity of theory is related to a failure to

appreciate the constraints placed on numerous urban activi-

ties by strong technical, marketing and other interrela-

tions. The integrating effect and pronounced nature of

these relations require us to consider the activities as

being bound together by an urban complex rather than sepa-

rate phenomena (Isard, 1960, pp. 673-4; Webber, 1967,

p. 401; J. M. Thompson, 1969),

Spatial activities are an excellent example. A con-

siderable portion of urban spatial activities are related

to the purchasing of goods and services. Consumer travel

movements are the result of urban spatial patterns as well

as agents affecting spatial structure (Huff, 1960; Webber,

1967, p. 401; Lowry, 1970). Thus consumer behavior should

represent an important concern in urban environmental design

both in terms of the location of shopping facilities and

the provision of access to these facilities. Knowledge of

the underlying determinants of consumer spatial behavior is

an essential input in the rational design of urban environ-

ments (Goldstucker, 19661 R. M. Downs, 1970, p. 13).

Marketing, along with other disciplines (McClelland,

1962, p. 133; Gutman, 1966, p. 114; Studer and Stea, 1966,

p. 128), must contribute this knowledge to the relevant
stages in the urban planning process (latridis, 1966,

pp. 473-83; Fagin, 1967; Fagin and Tarr, 1967) in order to

develop the proper synthesis of the five basic ekistic units

of nature, man, society, shells (building units), and

The complete urban planning process involves five
stages of (1) analysis of human needs and objectives and
general alternative courses of action, (2) establishment
of policy in which alternatives are ranked, on the basis
of local, regional, and national constraints and short and
long-term effects of psychological, social, economic, politi-
cal, and legal nature, and selection is based on probability
of satisfying present and future needs, (3) programming or
translating policy into a detailed course of action, (4)
design or planning of appropriate technical solutions and
(5) implementation or (a) organization and supervision of
execution of the plan, (b) programming for the adjustment
of the population to the new or altered physical environ-
ment, and (c) establishment of feedback mechanisms in order
to assess results, introduce further changes, and possibly
train residents.

networks (involved in the movement of man, goods, power,

and information) (Doxiadis, 1969). While motivated by

self-interest, this contribution is necessitated by con-

tinued malfunctions in the urban man-environment system.

Urban Planning and Environmental Malfunction

These malfunctions stem from urban planners' use of

unrealistic models, limiting terminology, inappropriate

research orientations, and "objective" behavioral criteria.

Cities are being patterned as planned areas bound

together by an arterial network with a sense of relatedness

derived from an aerial view. The resultant culturally uni-

form physical space (Hall, 1970, p. 17) is the product of

popular knowledge and rules of thumb for the design of

physical structures, separate land uses, traffic distribu-

tion functions, and urban growth. Such rules are couched

in fragmentary, static views of the city or the occasional

consideration of urban utopia (Lynch and Rodwin, 1958;

Webber, 1967, p. 403; Sommer, 1969, pp. 145-54).

Relationships in urban plans are typically created

within physical, rather than socioeconomic, space and are

distributed by means of a physical Euclidean spatial system,

rather than by non-Euclidean mapping. Such plans seem

designed for residents suspended at ten thousand feet and

content to spend their lives looking down. As a model of

spatial relations, a map plan implies an underlying theory.

The choice of a proper model should be an empirical problem

and thus correspond to the phenomena it is designed to

represent (Earvey, 1969a, p. 376).

Malfunctions are magnified by the employment of

historically preconditioned spatial designations (strip

center, department store, rest area, etc.) which focus on

physical characteristics and prematurely close space by

representing solutions rather than statements of environ-

mental problems (Kates and Wohlwill, 1966, p. 19, Michelson,

1970, pp. 50-31 Studer, 1970, p. 58). Such errors arise

from confusing the assessment of designed environments in

terms of their influence on behavior with the study of

extant environments and their impact on behavior. In the

former case behavior constitutes a class of independent

variables while in the latter case the opposite is true

(Gutman, 1966, pp. 104-51 Studer and Stea, 1966, pp. 130-3).

Urban malfunction is the result of isolating physical

planning from human needs and either ignoring or failing

to initially specify requisite behavior (Duhl, 1963, p. 151,

Studer, 1970, p. 58). A realistic urban model contains

elements which are functionally related to user paths and
action spaces (Wolpert, 19651 Kates, 1966, p. 261 Carr and

3Both the poor planning of urban designers and the
lack of planning by individual entrepreneurs contribute to
environmental malfunction. Thus knowledge concerning cus-
tomer behavior can provide valuable insights for both urban
planning and marketing strategy.
4An individual's action space is that perceptual and
objective area with which he has contact and within which
his spatial activities (economic, social, political, reli-
gious, etc.) take place.

Schlissler, 1969; Horton and Reynolds, 1969; Appleyard,

1970) and focuses on the effective environment (Gans,

1968, p. 61), that version of the potential environment

that is manifestly or latently adopted by users. Persist-

ent disparities between perceived and officially designated

areas are the natural consequence of outdated and unrealis-

tic urban models. Although the automobile and mass trans-

portation have replaced the horse, parking facilities

reflect little notice of technological change. The traffic

artery and curb parking have replaced the shopping promenade

without necessary alterations to reconcile the incompatibil-

ity of through traffic and ready ingress and egress (Gallion

and Eisner, 1963, p. 218).

Sources of urban malfunction have been attacked under

the functionalist precept that "form follows function" but

such efforts have reduced themselves to urban expressionism

(Studer and Stea, 1966, p. 129). More significant has been

the use of explicit behavioral criteria in developing plan-

ning models (latridis, 1966). Unfortunately, the behavioral

functions which have been adopted typically correspond only

to an objective urban physical pattern. Thus least effort

principles are seen as governing consumer spatial behavior

and convenience has become a chief consideration in planning

amenities (Parr, 1966, p. 42; Willis, 1969, p. 189). There

is a pressing need for urban plans which begin with a defini-

tion of empirically determined requisite behavior and environ-

mental needs rather than the specification of a physical


Behavioral Science Contributions

The behavioral sciences offer the most obvious route

for the effective translation of relevant behavioral goals

into urban plans or plans for the provision or retail

facilities in an urban complex. Yet the actual contribu-

tion of the behavioral sciences has been limited by the

availability and character of past research (Rosenberg,

1965, p. 3).

Anthropologists have tended to view the environment

as either a setting for primitive culture or as a single

physical setting over which culture has triumphed (Kates,

1966, p, 22; Schnore and Lampard, 1967, pp. 29-32). Soci-

ology and psychology, with few exceptions (Heyman, 1964;

Kates and Wohlwill, 1966; Craik, 1968; Michelson, 1968,

1970; Proshansky, Ittelson, and Rivlin. 1970; Wohlwill, 1970),

have adopted an implied doctrine of environmental neutrality.

Sociological conceptualizations picture the environ-

ment as a dependent function of social organization and

change (Kates, 1966, p. 22; Schnore and Lampard, 1967,

pp. 29-32). Even human ecology, which studies the relations

between the physical environment and social science varia-

bles, treats space as a medium rather than a variable with

a potential effect of its own. The tendency to concentrate

on aggregate behavior has prevented ecologists from dealing

with the more microscopic relationship between the individual

or household and the environment. Human ecology is rooted

in a period in which the major consideration was human adap-

tation rather than active intervention to create an environment

capable of satisfying a multitude of explicit goals. Such

intervention creates demands for a new variety of knowl-

edge (Michelson, 1970, pp. 17-22).

Lacking such information, sociologists have shared in

the planning process by substituting their own behavioral

and social criteria for those of future residents. They

have neglected even the most apparent trade-offs, e.g. con-

venience versus density, and blindly adopted maximum social

intercourse as the principle goal of planning. Urban

sociological research tends to emphasize residential units

while excluding the total pattern of relationships which

help shape human and consumer spatial behavior (Willis, 1969,

p. 189).

Psychologists traditionally accept the biotic and

physical environment dichotomy and concern themselves

almost exclusively with biotic relationships. Studies of

perception and symbolization of discrete stimuli either

ignore the urban environment or treated it as neutral (Som-

mer, 1966, p. 59; Wohlwill, 1966).

In the behavioral sciences "the proper study of man-

kind has been man" (Kates, 1966, p. 22). The physical

environment has been the research domain of the health and

design related professions. Thus environmental stimuli are

presented as specific etiological agents of sickness, com-

plex systems relevant mainly to lower life forms, or objec-

tive physical structures (Kates, 1966, p. 24).

The small body of behavioral research which does

focus on environmental matters must be approached with

caution. Attempts to transfer findings based on study of

lower animal life forms tend to confuse analogy with homol-

ogy. Laboratory research usually avoids the degree of

complexity present in the range of variables otherwise

studied in an urban setting. Even those studies using

multivariate designs outside the laboratory generally ignore

the cityscape (D. L. Thompson, 1966; Griffin, 1969; Green-

bie, 1971). Finally, it is doubtful whether environmental

concepts developed at the level of the room, dwelling,

neighborhood, etc. (or shop, department store, center, etc.)

are applicable to the urban scale or whether "scale free"

concepts are possible to develop (Kates and Wohlwill, 1966,

p. 17).

Although the desire for a more rigorous theory of

urban structure has led behavioral and social scientists

to study the dynamics of urban retail structure, they have

generally neglected customer behavior (Scott, 1970, p. 168).

Thus, in spite of an existing body of environmental and

urban behavioral research and a variety of analytic tech-

niques, there is an acute absence of both basic research

and coherent theory dealing with the retail structure as

an urban environmental component (Scott, 1970, p. 181).

The Need for a Model of Customer Spatial Behavior

The development of a model of customer spatial behavior

is thus a requirement for progress in forming a theory of

urban structure and reducing the malfunctions inherent in

urban plans. Armed with such a formulation, marketing can

both contribute to the urban planning process and help pre-

dict the consequences of alternative urban plans. Similarly,

such a model would help marketing management forecast the

consequences of changing urban structure. Thus there is a

need to "become more theoretical in order to become more

practical" (Alderson, 1965, P. 4).

At the theoretical level, a conceptual framework will

provide an alternative to mere fact gathering, help estab-

lish the relevancy of collected data, and provide a frame

of reference for future research (Kornhauser and Lazarsfeld,

1964, p. 15). It would allow research to be integrated into

a meaningful whole and provide a perspective for assessing

the significance of new research. The model would be use-

ful for consumer and urban theory construction and for the

development of testable hypotheses. Finally, it would con-

tribute to our knowledge concerning the performance of con-

sumers' psychological systems (Engel, Kollat, and Blackwell,

1968, pp. 35-6).

Customer Spatial Behavior, Environmental Behavior, and
the Requirements for a Model of Environmental Behavior

To develop such a body of theory we must begin by

drawing a conceptual distinction between the human behavior

and its subsystem, environmental behavior. The environ-

mental behavior subsystem involves "the kind of human action

which addresses itself to space, place, or environment as a

whole, not simply to other human actors" (Blant, 1969, p. 8).

The relationship between customers and a proposed or exist-

ing urban retail structure has a similar spatial dimension

which is a special case of environmental behavior and thus

distinct from other types of consumer behavior. To recog-

nize this relationship is to form the basis for a link

between the customer and the urban physical matrix in a man-

ner not unlike the link which others have sought to estab-

lish with respect to the human matrix (Kover, 1967).

To facilitate the study of customer spatial behavior

we must place it within the context of an environmental

behavior model. While retaining the integrity of a distinct

and special case, customer spatial or environmental behavior

will thus share the essential characteristics of the more

general subsystem. Since the nature of environmental behav-

ior rarely has been explored, it is necessary to begin by

setting forth, in general terms, the requirements for a model

of such behavior. These criteria include:

1. The description and analysis of environmental
behavior must be ego-centered.

2. The relationship between a? individual and
an environmental situation) must be seen as a
reciprocity, involving a dual interchange.

5In the general case the term environment refers to
the physical environment as opposed to non-physical cul-
tural, social, and psychological systems. The present

3. The environmental situation must be seen as
possessing objective and phenomenological

4. The environmental situation must be seen as
influencing environmental behavior by acting
as a behavioral context and/or as a proximal
or distal stimulus.

5. The model of environmental behavior includes
four basic elements which must be related to
one another in terms of intersystem congru-

It is necessary to examine these criteria and their

implications and determine how they influence the framework

of a model of consumer spatial behavior. The model would

seem to be suitable if its elements and relationships are

conceptually consistent with the general perspectives of

theory and practice in marketing and urban planning, help

correct present misconceptions concerning urban spatial

behavior, and provide a realistic explanation of the phenom-

ena under investigation.

Ego-Centered Environmental Behavior

It is necessary for any model building process to

include an appropriate conceptual scheme which resembles,

in certain desired features, the phenomena under investi-

gation (Rigby, 1965, p. 112). Thus at the outset attention

concern, however, is confined to urban settings. Hence
environment refers to the man-altered urban physical envir-
onment. The term does not imply a denial of the role of
important urban physical elements which are not man-made.
Instead, the man-altered physical environment represents a
special set of cases along urban to rural and natural to
man-influenced continue (see Craik, 1968).

must be given to the unit of observation (Chapin, 1961,

p. 305).

The ego-centered criteria does not deny the value of

studying spatial systems, cultural regions, group environ-

mental behavior, group spatial needs, or macroscopic con-

nections. Instead it casts the model at the lowest reduc-

tive level capable of producing micro-theory in order to

relate the environmental situation to the smallest unit

common to all social levels, the individual (Kain, 1967,

pp. 163-4; Blaut, 1969, pp. 8-9; Michelson, 1970, pp. 45-6).

Thus, as an active environmental participant, the individual

becomes an integral part of the problem (Fatouros, 1968,

p. 520; Proshansky, Ittelson, and Rivlin, 1970a, p. 4). In

this manner space preference (Isard, 1956, pp. 84-5, 144-5;

Huff, 1960), or any other exogenous social or psychological

variable, can be seen as operating in an environmental

situation and information field and accounting for environ-

mental response variation. The identification of patterns

of environmental response becomes possible because such

behavior is enduring and consistent over time and situation.

Yet only an ego-centered view can cope with the expected

diversity through time and space, and with the continuous

variability in any given space and time, of environmental

response behavior (Proshansky, Ittelson, and Rivlin, 1970c,

pp. 29-31).

Such variance has been examined in terms of larger

units of observation (Huff, 1962b, 1963, 1964) but such

efforts have not proven successful. Alternatively, an

implied belief in the adaptability of economic institutions

has led to a total concern with urban retail structure as

a guide to customer spatial patterns. Yet, even if allow-

ance is made for this assumption, it is not stores that

visit one another. The spatial character of environmental

behavior arises for the fact that people are separated from

each other and from non-residential activities (Michelson,

1970, p. 46).

The distinction between consumer and human behavior

is a necessary artifact. The absence of an adequate model

of human behavior has not abated the defense of separate

models of consumer behavior. These formulations have been

defended on grounds of reported differences in the consumer

role as opposed to other types of behavior and, ironically,

on the basis of the synonomous nature of human and consumer

behavior as evidenced by the pervasive association of con-

sumption goods and all human activity (Engel, Kollat, and

Blackwell, 1968, p. 5).

In truth, such distinctions are a matter of theoreti-

cal perspective (Sommers and Kernan, 1966, p. 5). The

productivity of the role distinction and the absence of a

general theory of human behavior form the best argument for

the creation of separate models. To the extent that all

behavior is choice behavior or that consumer and human

behavior are identical, studies in the former enrich our

general understanding (Tucker, 1966, pp. 55-60). Instead

of precluding the employment of ego-centered customer

models, this realization limits the character of analysis.

It requires careful discrimination in borrowings from other

disciplines dealing in other roles and/or other behavioral

subsystems and the simultaneous development of concepts,

tools, and hypotheses suited to the peculiar nature of cus-

tomer environmental behavior (Halbert and Durkson, 1966,

p. 30; Myers and Reynolds, 1967, p. 2).

This approach is consistent with marketing theory

which regards the consumer as the central factor in the

marketing process. While not limited to the study of the

consumer, marketing theory is advanced by increments in our

knowledge concerning customer behavior. Such advances are

the basis for improvements in marketing practice and the

economic system, whose success depends on the ability of

the marketing process to anticipate, create, and deliver

satisfactory assortments (Alderson, 1965, p. 144; Nicosia,

1966, p. 51 Sommers and Kernan, 1966, p. 3). The ego-

centered criterion can furnish a basis for added insight

as long as tendencies to universalize findings are couched

in the realization that the urban context is merely one case

and customer spatial behavior is merely one facet of con-

sumer behavior.

Microscopic spatial theory has been labeled as a bar-

rier to the "science of society as a whole" (Stewart, 1948,

pp. 31-2) because it implies an examination of properties

which are not easily observed. Larger units of observation

involve less demanding methodology but their value as plan-

ning units is suspect.

Thus, the small group, a convenient empirical unit,

tends to be associated with severe planning problems when

its dynamics are generalized to a higher urban scale (Rosen-

berg, 1965-6, p. 5). The neighborhood, a useful sociologi-

cal concept, has a subjective quality which prevents its

ready manipulation in an urban model (Willis, 1969; T. Lee,

1968). Yet larger residential units, which typically com-

prise the divisions of urban plans, do not correspond to

urban sociological structure. It is difficult to make

observations at this level because these varied and over-

lapping residential areas are perceived differently by


Theoretical and empirical considerations and current

trends in social science research have made the household

the most serious alternative to the individual as a primary

unit of behavioral analysis. This view holds that certain

basic human needs are "met only in the family, making it

a small, unitary social system, having its own independence

and integrity" (Chapin, 1961, p. 306). Thus it is main-

tained that consumption related behavior occurs in or through

a household unit which possesses unique properties that act

as mediating variables but significantly differ from the

functionally analogous intervening variables which charac-

terize individual behavior (Sharp and Mott, 1956; Wolgast,

1958; Coulson, 1966). The notion of the customer purchasing

agent and household representative entering the market to

replenish or extend an assortment of goods needed to sup-

port expected patterns of future behavior has been reinforced

by evidence indicating the role of the family in purchasing

decisions (Foote, 1961, p. 5; Whiteside, 1964; Alderson,

1965, pp. 144-5; Myers and Reynolds, 1967, pp. 245-6, 261).

Close observation, however, has revealed this agent to

be an imperfect interpreter of household preference. Such

preference tends to vary along product and brand dimensions

and constitutes only one of a configuration of variables

related to purchase intention (Coulson, 1966, pp. 1-2).

Furthermore, the typically vague demarcation between con-

sumer and customer roles suggests that a general category

of customer behavior be used to include both cases when the

two roles are congruent and when only pure customer behav-

ior is involved (Halbert, 1966, p. 72). Thus environmental

behavior related to the retail structure is designated cus-

tomer environmental behavior. In this way the integrity of

the ego-centered unit is preserved by allowing the previ-

ously considered variables to be regarded as factors affect-

ing the individual decision process.

In effect this rules out the application of methods

designed to apply to aggregate behavior. Included in this

category is a set of tools based on the gravity concept of

interaction. Although they have been variously modified to

account for spatial response to urban retail structure

(Carrothers, 1956; Gist, 1968, pp. 164-73; Scott, 1970,


pp. 169-78), gravity models are essentially empirical tools

useful only for macroscopic description. They represent

spatial analogies to Newtonian physics which have grown out

of a dissatisfaction with the descriptive progress of micro-

scopic theory and the conviction that group members exhibit

different behavior than they do as individuals (Stewart,

1947, 1948, pp. 31-2; Isard, 1960, p. 464).

The gravity concept postulates "that an attracting

force of interaction between two areas of human activity is

created by the population masses of the two areas, and a

friction against the interaction is caused by the interven-

ing space over which the interaction must take place" (Car-

rothers, 1956, p. 94). This space, urban or otherwise, is

conceived in Newtonian terms as a mass structured according to

principles governing "in an overall fashion the range of

behavior of individual particles, both constraining and

initiating their action" (Isard, 1960, p. 494). The restric-

tion to population and distance, although they may be con-

sciously employed as general indexes of attraction and repul-

sion (Gist, 1968, p. 168), and the application to human

aggregates lead to the assumption that the idiosyncracies

of individual or even small group behavior are cancelled or

averaged out (Isard, 1960, pp. 494, 512-5).

This idea conflicts with the ego-centered proposition

and contradicts evidence that both individual customer

behavior and the effect of the parameters of spatial inter-

action are significantly different (Huff, 1962a, pp. 64-5;

Moore and Mason, 1963; Revzan, 1966; Scott, 1970, p. 170;

Mason and Moore, 1970-1). Unfortunately, while disaggre-

gation and stratification of gravity data lend additional

precision to findings concerning spatial behavior, such

efforts often prove impossible in practice.

This restricts the role of gravity formulations to

macroscopic description (Isard, 1960, pp. 515-6; Moore and

Mayer, 1966, p. 830; Kain, 1967, pp. 163-4; Mason and Moore,

1970-1) and excludes their application as predictive plan-

ning tools. Attempts to employ variables which seem statis-

tically related to urban behavior at the aggregate level

have proven almost universally unsuccessful at reduced levels

(Stewart, 1942; Blount, 1964; Horton and Reynolds, 1969,

p. 71). The failure of gravity models is merely one example.

Reciprocity and the Customer-Environment Interchange

The order in spatial behavior, which is sometimes

exposed but never explained by aggregate concepts, is the

product of environmental interaction. In this context it

is hardly surprising when observed empirical regularities

are absent when similar retail interaction phenomena are

studied under differing environmental circumstances (Huff,

1962a, pp. 66-7).

The dynamics of the customer-environment interchange

suggest that plans for the retail structure and the total

urban complex reflect a knowledge of human spatial needs,

uses of space, and the social context of these uses, as well

as the economics of spatial usage because the ultimate

product of urban design is a physical matrix within which

a social system will operate (Proshansky, Ittelson, and

Rivlin, 1970a, p. 493). A customer does not merely move

through a neutral environment in order to satisfy his needs,

search for variety, etc. His relationship to that environ-

ment is an active unity involving both the satisfaction of

needs and the influence of the environment on behavior.

This interaction results in the transformation of the indi-

vidual and may create new needs or other behavior modifica-

tions. These altered characteristics interact with the

environment to produce changes in the latter, thus creating

a continuously changing mechanism. The redefined reciprocity

between the customer and the urban environment may be con-

sidered a state of learning in a continuous learning process.

The changing individual condition and environmental situa-

tion suggest that the environment is a field of learning

and that the customer's active participation through envir-

onmental contact is an involvement of learning (Rosenberg,

1966, p. 4; Fatouros, 1968, p. 520; Proshansky, Ittelson,

and Rivlin, 1970c, pp. 34-5).

This effectively excludes the isolated study of either

human or physical variables. Instead, the analyses of this

active unity should proceed by conceptualizing interaction

as existing primarily with respect to the structure of the

urban matrix rather than with its individual parts or ele-

ments. In this way the field of learning is viewed as a

set of relations between environmental elements (Fatouros,

1968, p. 520).

This perspective challenges the theoretical founda-

tions of the gravity concept and casts doubt on the poten-

tial contribution to a theory of consumer environmental

behavior of goods classification systems, trade area deline-

ation procedures, and decision process models.

Unlike its Newtonian counterpart, the gravity concept

does not include all relevant variables known to be opera-

tive in the real world. This neglect of the interrelated-

ness of urban activities acts as a detriment to the analysis

of the urban system. As its most important consequence,

this means that gravity models are neither theoretically

nor empirically suited for urban retail application (Huff,

1962a; D. L. Thompson, 1966, pp. 5-7).

By reducing environmental factors to population and

distance measures, gravity models neglect the elements and

relations present in urban space and effectively eliminate

the urban environment as an operating influence. The dis-

tance concept, for example, fails to comprehend the dif-

ferential character of urban space, a function of such vari-

ables as travel mode, store and good type, community size,

topographic conditions, population density, work place, and

retail agglomeration and diversification. Further, it

ignores those ego-centered variables which are associated

with different spatial habits and tendencies (Goldstucker,

1966, pp. 299-300; Moore and Mason, 1966; Scott, 1970,

pp. 60-5). While gravity advocates have admitted the rela-

tion between the nature of urban structure and variance in

spatial behavior (Converse, 1949, p. 383), gravity models

have been modified only substituting other variables for

population and distance and applying them to different types

of urban competitive conditions (Gist, 1968, pp. 171-2).

But simple summary variables are neither suited for

spatial analysis of urban retail competition nor capable

of dealing with relevant environmental influences or their

urban character. These defects have left theorists in con-

fusion as to the non-uniformity of time and distance in

urban settings. Hypotheses functionally relating distance

to population or itself have attempted to resolve this

debate and have revealed an implicit, but insufficient,

recognition of urban environmental influence (Carrothers,

1956, p. 97; Scott, 1970, p. 170).

Such limitations can be traced to the essentially rural

character of the gravity concept. Reinforced by apparent

support for the analogy between rural population and gas

molecules, gravity advocates have sought to extend the com-

parison to an urban context. This has been accomplished by

either ignoring environmental differences or by regarding

population density as a process of molecular bonding and the

city, a high density gas, as a liquid (Stewart, 1948, p. 54).

Yet despite urban retail applications (Ellwood, 1954; R. L.

Nelson, 1958, p. 149; Lakshamanan and Hansen, 1965; Voorhees

and Lakshamanan, 1966), the gravity concept remains dependent

on the neutrality of intervening space. Its origin as a

concept applicable to cities of some size and a number of

miles apart has prevented its successful use in the analysis

of customer movement within the limited confines of a city

containing a broad assortment of goods (Reilly, 1929,

pp. 13-15; Converse, 1949, pp. 382-3; Huff, 1962a, D. L.
Thompson, 1966, p. 5).

Customer response to scrambled merchandising, retail-

ers stocking more than one type or class of product, and

the existence of different type or even multiple purpose

shopping trips serve to destroy the gravity concept as a

vehicle of urban analysis (Huff, 1964, p. 361 Sparicio, 1966).

In retrospect, the limited applicability of gravity models

to nearby populations in studies involving a broad geographic

frame (Stewart, 1965, pp. 70-1) seems to be an early indi-

cation of the eventual failure of intra-urban trade are

delineation attempts (Reynolds, 1953; Mayer, Johnson, and

Mason, 1970).

One solution has been the recharacterization of urban

space in terms of decision alternatives. Thus, the inter-

vening opportunities model denies any necessary relationship

between distance and movement ard posits the number of indi-

viduals traveling a given distance as being directly propor-

tional to the number of opportunities at that distance and

inversely proportional to the number of intervening opportu-

nities (Stouffer, 1940). Retail applications of the model

have served to indicate that distance fails to perform a

continuous function and thus have provided a conceptual

foundation for the probabilistic definition of demand sur-

faces (Huff, 1963, 1964).

The value of redefined urban space is, however, over-

shadowed by difficulties in defining opportunities, reliance

on inadequate concepts of behavioral equilibrium, dependence

on largely homogeneous opportunities and population and a

neglect for the variable influence of distance (Scott, 1970,

pp. 62, 179-80). While the opportunity concept treats space

in a manner generally consistent with the distribution of

retail alternatives, it still implies that the intervening

urban space is essentially neutral.

The ultimate source of this confusion is the rural

orientation of retail theory. Although retail establish-

ments and sales have been historically concentrated in met-

ropolitan areas, the earliest analysis of such commerce

focused on the distribution of retail trade in rural areas.

While rural patterns were originally investigated because

they involved a more manageable unit of analysis, the con-

clusions drawn from these studies have formed the basis for

current views on consumer shopping behavior and trade area

analysis. Current views of patronage distribution, for

example, arise from the marked overlapping and differential

configurations observed in early delineations of social

zones of retail influence which revealed different travel

patterns associated with fashion and shopping goods as com-

pared to convenience or bulky items. The complex spatial

distribution of demand in the rural commercial order

implied the operation of a "highly selective mechanism of

consumer choice" (D. L. Thompson, 1966, p. 2).

Differences in urban and rural spatial habits and

attitudes (Swedner, 1962; Goldstein, 1966; D. L. Thompson,

1966) suggest that a different and perhaps more complex

mechanism may operate in the urban context. But such dis-

tinctions have been ignored and efforts have been devoted

to identifying customer spatial regularities in trade are

studies which relate travel to types of goods and their

associated distributions (W. A. V. Clark, 1968, p. 396;

Scott, 1970). The development of the current classification

of consumer goods (Copeland. 1923; McCarthy, 1971. pp. 300-19)

(convenience, shopping and specialty) and the concept of

range (Berry and Garrison, 1958a, 1958b; Garner, 1967;

Sparicio, 1966; Scott, 1970, pp. 13-15, 157-66) are the

most significant developments in this theoretical tradition.

Such constructs are unjustified abstractions which

neglect the differential influence of the total urban

environment. Relationships drawn between individual prod-

ucts and customer spatial tendencies ignore the existence

and influence of retail institutions (Jacobi and Walters,

1958; Bucklin, 1963). This produces numerous empirical

oversights and prevents the systematic treatment of such

factors as the coincidence of search and purchasing trips

or multiple purpose trips in which different types of goods

are purchased together (Murdie, 1965, p. 212; Kleimenhagen,

1966-7, p. 39; W. A. V. Clark, 1968, p. 387).

The absence of unique product classifications over

time and the simultaneous assignment of multiple classifi-

cations to the same physical product provide the realiza-

tion that the conventional goods classification system is

actually a device for the discrete designation of spatial

buying patterns. Such spatial responses are observed with

an implied model consisting of three variables, consumers,

a distribution of goods, and marketing mix influences. The

retail structure itself plays a passive role. Certain con-

tradictions between the spatial principles advanced by the

goods classification system and empirical evidence are

rationalized by the more recent assortment concept (Alder-

son, 1965, pp. 144-9) insofar as it implies a multiple pur-

chase behavior tendency. Still the level of abstraction

continues to prohibit a systematic treatment of urban retail

relations and other environmental variables.

The absence of environmental considerations has also

limited the character and value of trade area analysis and

retail location models as predictive techniques designed to

deal with the provision and location of retail facilities

and sources for the development of a theory of urban struc-

ture (Scott, 1970, p. 168). Such investigation was origin-

ally limited to estimates of population within an arbitrarily

determined market area (Imus, 1961) with occasional analysis

of composition by income group. A variety of ratios of

frontage or floor space or number of stores to population

were employed in this connection. More recently, urban

population shifts, metropolitan growth, rising per capital

income and attendent changes in demand and car ownership,

and changes in the technology and structure of retailing

have increased both the size and risk of urban retail invest-

ment and necessitated more reliable analysis (Scott, 1970,

pp. 171-2).

The result has been the development of a loosely con-

nected five-stage evaluation process consisting of trade

area delineation, determination of gross and net adequacy

(potential), forecast of expenditures for relevant products,

and site evaluation (Gist, 1968, pp. 161-228). While envir-

onmental variables are considered in the final stage, their

presentation is typically limited to descriptive models of

retail structure and loosely related "principles" of site

evaluation which either rely on social physics notions or

focus on physical relationships and neglect reciprocity as

a spatial determinant of demand (Cohen and Applebaum, 19601

Gist, 1968, pp. 201-281 Scott, 1970, p. 18).

The omission of reciprocity has consequences which

extend beyond notions of gravity, goods classification or

trading area. The most important effect is the absence of

environmental response in models of consumer decision pro-

cesses (Nicosia, 19661 Engel, Kollat, and Blackwell, 1968).

In fact, such models remain trapped in the rural abstractions

of earlier formulations and lack a comprehensive descrip-

tion of the morphology and mechanisms associated with

store choice.

The relationship between the customer and the retail

structure is defined in terms of a purchase process (Nicosia,

1966, pp. 179-84; Engel, Kollat and Blackwell, 1968, pp. 444-

50). Actually this stage is limited to customer-store

environment interaction and thus exemplifies the neutrality

doctrine with respect to variables external to the store

environment, except where, like parking facilities, they may

be treated as store characteristics.

Preceeded by problem recognition and possible stages

of external search for alternatives and evaluation of alter-

natives, the purchase process involves four related sets of

variables, a prepurchase intentions typology, consumer char-

acteristics, store environment characteristics, and purchase

outcomes. With consumer and store characteristics acting

as intervening variables, the remaining factors are employed

to form an empirically identifiable intentions-outcome matrix

along product and brand dimensions. The result is a severely

restricted role for environmental factors and a failure to

integrate store choice and purchase processes.

Store choice (Engel, Kollat and Blackwell, 1968,

pp. 451-64) is never actually integrated into decision

process theory. It exists instead within a largely inde-

pendent scheme composed of four variables, including evalu-

ative criteria, perceived characteristics of stores, a

comparison process, and an acceptable-unacceptable store

dichotomy. The fixation on product and brand choice results

in this scheme containing fewer explicit relationships than

might be expected. The nature of the comparison process,

for example, is never made clear. It is not specified

whether a direct comparison between perceived store char-

acteristics and evaluative criteria, or an indirect compari-

son, involving surrogate indicators and/or standardized store

choice strategies occurs. Further, the tendency to focus

almost solely on product and brands has left the mechanism

producing the acceptable-unacceptable store decision unex-


Yet the consequent rationalization holds, even in the

absence of published evidence, that the bulk of patronage

is not dependent on the store choice process but instead is

usually the result of satisfactory experience (Engel, Kollat

and Blackwell, 1968, p. 451). The conditions necessary for

the absence of the store choice process are stated as unknown.

Thus in the final analysis, decision process models follow

the doctrine of environmental neutralism. While, in many

respects, they represent a synthesis of the most recent

advances in the theory of consumer behavior, they are incap-

able of significantly contributing to a model of environ-

mental behavior and thus to the theory and practice of urban


The reciprocity proposition requires a reliance on

constructs which treat spatial behavior and the retail

structure as interdependent. Store choice, thus becomes

the process, rather than the act, yielded from the interac-

tion of changing customer characteristics and changes in

the structure of the urban matrix and its retail component.

The limited contributions of current approaches suggests

the need to establish a new framework for such analysis.

Dual Definition of the Environmental Situation

Environmental definition has ranged from objective

physical description to phenomenological characterization.

Neither extreme is desirable for our purposes. In the

former case the environment is fragmented into discrete

quantifiable stimuli which are specifically related to

experience and behavior in a manner consistent with experi-

mental and Watsonian behaviorism. The phenomenological view

denies the validity of this position and instead treats the

environment as it is experienced, in terms of transformed

symbolic stimuli which comprise an inner psychological

environment (Kates, 19661 Proshansky, Ittelson, and Rivlin,

1970c, p. 28).

Philosophic antecedents of objective definitions may

be found in Laplacian optimism which maintained the concep-

tual possibility of describing the history and future state

of each atom in the universe. Environmental determinism,

an outgrowth of simplified Laplacian idealism, transformed

the original contention into a belief that human behavior

was predictable in terms of the environment. The development

of the Heisenberg uncertainty principle and the growing

conviction that environmental response variation was some-

how related to environmental perception however, cast doubt

on such simple cause and effect interpretations (Hagget,

1966; H. M. Mayer, 1967, p. 231; Mann and Hagevik, 1971).

The resultant debate between advocates of determinancy

and response variation, rooted in some form of free will,

has produced numerous compromises such as variations of

satisficer models. Most significant, however, has been the

development of probability constructs which treat the oppos-

ing views as a matter of scale (Hagget, 1966, pp. 23-6).

Elements of both forms of compromise are found in Huff's

(Huff, 1962b, 1964) approach to consumer spatial behavior.

Dropping the requirements for perfect choice discrimina-

tion and maximizing behavior, Huff modified earlier gravity

formulations in order to calculate the relative expected

utilities of different retail alternatives. Relative utili-

ties are then used to explain patronage distribution in

probabalistic terms and account for observed variance in

shopping patterns. The chief flaw in the Huff model is

limiting shopping alternatives to those within a given

radius. The model assumes away a critical factor in the

choice process and produces probability statements which

conflict with observed dimensions of habit, inertia, and

problem solving (D. L. Thompson, 1966, p. 8). An analysis

of these weaknesses provides an insight into an improved

method of defining a given environmental situation.

The ego-centered study of environmental behavior sug-

gests, as a first approximation, a dichotomy involving the

customer participant and the environmental situation. The

existence of a single set of physical environmental condi-

tions does not, however, result in single, uniform view of

the situation (Proshansky, Ittelson, and Rivlin, 1970c,

pp. 35-6). A multiplicity of views are produced by the

active unity of reciprocity, the phenomena of selective

perception, and differences in perception.

The first two characteristics of a model of environ-

mental behavior imply that the customer does not passively

react to environmental factors impinging on him. A variety

of cognitive activities, including expectancies, attitudes,

and symbolic transformations mediate and modulate the impact

of environmental forces (Kates and Wohlwill, 19661 Sonnen-

feld, 1966). Complete attention to the physical effects of

the urban environment ignores the human propensity to sym-

bolize experience and to react to both physical and symbolic

stimuli (Kates, 1966, p. 23). The perception underlying a

customer's reaction to physical urban stimuli and their

associated symbols is the product of environmental experi-

ence as well as expectancies, built up with regard to

environmental constants to which he has become habituated.

To conceptualize the relevant environmental phenomena

to which the customer reacts, it is necessary to define the

environment in a manner which comprehends the relationship

between the physical and symbolic environment and environmental

response (Proshansky, Ittelson, and Rivlin, 1970c, p. 28).

This requires the use of a dual definition which embodies

objective and phenomenological facets.

This position must, however, be immediately qualified

to account for selective tendencies. Since the urban

environment contains more information at any moment than

the customer's cognitive capacity can handle, perception is

highly selective (Parr, 1966, p. 41; Carr, 1970, pp. 522-4).

Finite perceptual and storage capabilities influence the

character of customer decision processes (Wolpert, 1964,

p. 537; Parr, 1966, p. 41) and thus the nature of environ-

mental definition. The measure of visual or symbolic input

is not the number of different images or environmental ele-

ments which are present or even seen. Environmental defi-

nition continues to have objective and phenomenological

features but is limited to that which is relevant from the

ego-centered view.

Yet environmental selectivity is not uniform. Increas-

ing evidence indicates the presence of significant differ-

ences in environmental and spatial perception by urban and

rural inhabitants and among urban residents of differing

socioeconomic backgrounds (Marble, 1959; Swedner, 1962;

Goldstein, 1966; Scott, 1970).

Under certain circumstances these differences appear

to dominate the physical aspects of location in determining

customer spatial patterns (Marble, 1959).

The failure to consider these perceptual dynamics

limits the Huff model, despite its other strengths, to the

treatment of situations defined in terms of environmental

determinism. The inclusion of all retail alternatives

within a given radius implies the perfect perception of the

urban environment in all its objective complexity. To the

extent that Huff's comparison mechanism may operate, it

surely applies only to elements in the retail landscape

which are within the customer's problem space. Arbitrary

physical determination of alternatives makes perception a

function of travel time and distance (D. L. Thompson, 1966,

pp. 8-9) and directly conflicts with the requirement for a

dual definition by adopting a position of environmental


The diversity of elements in the objective retail

structure reinforces the effect of heterogeneous percep-

tion in making a dual definition of the environmental situ-

ation necessary. The unequal generative capabilities of

similar types of retail units, generally ignored by the

gravity approach (Davidson and Doody, 1966, p. 121), sug-

gests that the properly defined environmental situation

exhibits irregular features produced by both objective

retail diversity and perceptual heterogeneity (Sparicio,

1966; Gruen and Gruen, 1967, p. 320; Mason and Moore,

1970-71, p. 37). This invalidates any simple notion relat-

ing consumers and products or brands in an abstract neutral

space and presents a difficult problem to marketing theo-


Two conventional solutions offer, at best, only par-

tial resolution of these difficulties.

First we find attempts to relate store classification

system to the goods typology (Bucklin, 1963; Kleimenhagen

and Stampfl, 1968). In this way it is possible to account

for temporal changes in either system and to make the envir-

onmental nature of the retail unit more explicit (see Krug-

man, 1960, and Winkel and Sasonoff, 1970 for a different

approach). Such an approach still fails in only partially

defining the environmental situation and thus being unable

to deal with the variable response to similar types of

establishments (Huff, 1964, p. 34; Revzan, 1966; Gist, 1968,

pp. 175-8).

A more sophisticated refinement is offered by redefin-

ing the nature of satisfaction so as to give it global

qualities extending beyond product evaluation to involve

the experience surrounding acquisition (Cardozo, 1965;

Newman, 1966, p. 218; Steiner, 1966, pp. 209-12). Shopping

thus becomes an environmental experience occurring in a

spatial loop (or other configuration) and beginning at some

origin, proceeding over some route by means of some trans-

port mode, continuing through retail elements, and returning

to the place of origin (or other place perceived as the

"end") (Martineau, 1958; Newman, 1966, p. 218). The trip

involves elements of purchase and search, although search

is not limited to the shopping experience and may even be

chiefly associated with other activities and thereby restrict

the extent of spatial behavior. Extensivity is thus a vari-

able of considerable importance and a function of the per-

ception of the buying process, percentage of the budget

involved, type of goods, and the differential advantage of

these goods and stores as well as the perceived results of

prior search (Udell, 1966; Bucklin, 1967; Engel, Kollat, and

Blackwell, 1968, pp. 378-421; Gist, 1968, pp. 157-9; Mason

and Mayer, 1970a).

The global satisfaction concept and its implications

are consistent with the reciprocity notion but must be

complemented by a treatment of costs. If shopping is an

environmental experience then patronage is also a function

of the cost of traversing space rather than a function of

intervening space itself. Gravity, trade area, and goods

classification theorists have treated this relationship in

somewhat different manners.

In gravity formulations the distance exponent has

been adjusted to account for trip type. This is intended

to reflect individual unwillingness to travel the same

distance for all types of trips. Alternatively, it has

been suggested that this exponent may be a variable func-

tion of distance itself. Thus the by-products of urban

growth, such as congestion parking, traffic, etc.. are seen

to result in the friction per unit of distance against

patronage interaction being greater in an urban area than

in less developed space (Converse, 1949, p. 383; Carrothers,

1956, p. 97). The general omission of such considerations

in conventional gravity approaches has led to a criticism

of their ability to predict intra-urban customer movements

(Huff, 1962a, pp. 64-5). Such deficiencies render gravity

model incapable of dealing with situations in which the

perceived costs of traveling to a larger and closer retail

alternative are greater than those associated with movement

to a smaller and more distant alternative. Yet even with

exponent modification, there is little more in the gravity

approach than a recognition of the existence of costs or

friction. No attention is given to their specific roles in

an environmental decision process.

Other theorists have dealt with such matters by modify-

ing the spationally dimensionalized economic theory of land

use determination. They have employed the idea of costs and

satisfaction as an index of environmental experience and a

basis for explaining retail structure in terms of customer

search propensities (Holton, 1958; Kelley, 1958; Cox, 1959;

D. L. Thompson, 1967). Thus various explanations of spatial

movement and patronage employ cost and satisfaction as com-

mon denominators of urban environmental experience. In

this way the spatial hypotheses which actually underlie

goods classification theory have been made explicit (Kleimen-

hagen, 1966-7).

From the standpoint of the dual definition criteria,

the central issue is the assumption that patronage decisions

involve the objective analysis of costs and satisfactions

and the use of some maximizing criterion (D. L. Thompson,

1966, p. 11). Several questions arise in this connection.

Attempts to list and classify the costs associated

with shopping (A. Downs, 1961; Bender, 1964) have served

primarily as indications of potential sources of environ-

mental irritation rather than as relevant cost functions.

The distinction between potential and salient costs is

usually overlooked. It is interesting to note the lack of

corresponding efforts associated with the study of satis-

faction. This is an understandable example of the social

science tradition which concentrates almost exclusively on

malfunctions in social and behavioral systems. However, in

this case there is the crucial implication that cost and

satisfaction are positive and negative counterparts in a

continuous function. Thus the absence of environmental

irritants (costs) would correspond to the presence of a

certain undefined quantum of either satisfaction or slack,

except when an unusual coping mechanism or a non-positive

stress-tolerance threshold operates (Wolpert, 1966). This

relationship is not obviously apparent and neither is the

validity of a single measure for environmental cost and


Such obstacles would be surmounted by the use of a

method for comparing costs and satisfactions existing with

respect to the same environmental source (e.g., varying

degrees of traffic congestion), different environmental

sources (e.g., varying degrees of traffic congestion and

perceived character of intervening neighborhoods), and dif-

ferent environmental and nonenvironmental sources (e.g.,

varying degrees of traffic congestion, the perceived charac-

ter of intervening neighborhoods, and expected monetary sav-

ings). These unresolved difficulties (Isard, 1960, p. 256;

Goldstucker, 1965, p. 315) have been considered only in

terms of criteria supposedly employed by customers in mak-

ing environmental decisions.

In the simplest approach, satisfaction exists as either

a constant or a variable limited to association with a prod-

uct. This yields the conclusion, associated with the tra-

ditional concept of range, that a customer will seek to

minimize the travel cost for the individual good (W. A. V.

Clark, 1968, p. 387). Other more sophisticated deductions

suggest that the customer maximizes the result of his clos-

est travel effort (W. A. V. Clark, 1968), maximizes the

total results of shopping by equating some variation of

marginal costs and satisfaction (Holton, 1958; Kelley, 1958;

A. Downs, 1961), or maximizes the "return" from a given bud-

get of time, effort, and money (Cox, 1959).

These conclusions fail to account for the complexities

introduced by internal search and prepurchase search activi-

ties. Although it forms the basis for modern goods clas-

sification and trade area theory, the assumption of comparison

and rational calculation is not supported empirically and,

in fact, is contradicted by a growing body of evidence

(Mueller and Katona, 1955; Holton, 1959; Luck, 1959; Cun-

ningham, 1961; Tate, 1961; Alderson and Sessions, 1962;

Udell, 1966; Feldman, 1967; Mason and Mayer, 1970a; Mason

and Moore, 1970-1).

The foundations for a theory of rational customer

calculation are weak. In a sense calculation is a mathe-

matical, rather than a behavioral, concept which fails to

distinguish between significantly different dimensions of

problem-solving behavior (Wilding, 1968, p. 145; Wilding and

Bauer, 1968). Certainly the existence of risk (Bauer, 1960)

casts doubt on the inherent superiority of maximizing strate-

gies over risk minimization or acceptance as the basis for

rational action (Myers and Reynolds, 1967, p. 101; Popielarz,

1967, p. 371; Barach, 1968, p. 141). Since no one customer

strategy has clearly been designated as rational in market-

ing theory, a number of different decision orientations

(Nehnevajsa, 1966) may be termed rational. Actually, the

tendency to assume customers to be rational reflects the

failure of marketing theory to reach a compromise between

economic and behavioral science viewpoints (Udell, 1964-51

Nicosia, 1966, p. 38). Since problem solving behavior is

a relatively rare phenomena and since any link between

rationality and maximization produces little more than a

tautology (Katona, 1953; Myers and Reynolds, 1967, p. 50)

i.e., it is impossible to present a situation contradicting

it, this path of inquiry seems conceptually sterile.

It would seem proper to conclude that the present

state of knowledge concerning the effect and nature of

the urban environment prohibits the treatment of customer

environmental experience solely in terms of simple costs

and satisfactions. Attention must be devoted to the char-

acter of environmental experience. The heterogeneity of

ego-centered and environmental variables suggests that this

experience be examined in terms of its objective and phenom-

enological character.

The adoption of such a dual definition of the relevant

environmental situation prevents the reciprocity criteria

from reducing a model of environmental behavior to a theory

of retail development and institutional change. It thus

becomes possible to speak of changes in environmental char-

acteristics which could otherwise only occur over time. This

prevents reciprocity from being transformed into a concept

of retail institutional adaptation. The changes in the

retail structure, referred to by the reciprocity concept,

are a potential source of institutional change but are them-

selves limited to modified perception. This aspect of reci-

procity presents an extension of the restructuring process

in a topographical model of consumer space preference (Huff,

1960, p. 164).

Contextual and Stimulus Properties of the Environment

The dynamic customer-environment interchange influ-

ences environmental response by providing a behavioral

context and stimuli which have either a direct impact or

generate long-term residual effects (Kates and Wohlwill,

1966; Wohlwill, 1970).

All behavior occurs within some physical context but

its objective and symbolic character, as suggested by the

dual definition, make that context far from obvious. The

customer himself may be unaware of the physical context

within which he acts. Thus the surroundings may appear

neutral and enter awareness only when they deviate from

some adaption level (Proshansky, Ittelson, and Rivlin, 1970c,

pp. 36-7). Yet studies involving so-called behavioral set-

tings, and including retail stores, reveal that certain kinds

of behavior are localized in certain types of settings.

People seem to learn to act in a particular manner in par-

ticular settings (Barker, 1963, 1968). In this limited

sense the environmental characteristics of a retail store

possess meaning as a behavioral context.

This, however, is too narrow a view of contextual prop-

erties. As noted previously it is dangerous to abstract

the retail environmental component from the relevant urban

environmental system (Proshansky, Ittelson, and Rivlin,

1970c0 p. 34). To do so is to invest the store environment

with contextual qualities which it does not possess and thus

to give the retail environment greater reality than the

environmental situation from which it has been abstracted.

The error of this procedure would be seen by the hypothetical

placement of identical retail units in different parts of

the same city or in different cities and then the observa-

tion of the behavior of the same customers in each context.

While such data is not available, existing information

does imply that customers relate to the total environment

that is personally relevant, including the environment sur-

rounding the behavioral setting of a retail element (Feld-

man and Star, 1971). Theoretically this is an example of

the proposition that the environmental setting which defines

a given behavioral context is not a closed system. Its

boundaries are not fixed in time or physical space (Proshansky,

Ittelson and Rivlin, 1970c, p. 31), but in the mind of the


A central feature of urban environmental research is

the effort to relate objective stimulus properties, their

perceptual manifestations, and the stimulus properties of

the symbolic environment. Such information is a prerequis-

ite for intelligent urban planning yet, to the extent that

the urban environment has been studied at all, only certain

objective stimulus properties have been traditionally

examined (Kates, 1966, pp. 23-4; Kates and Wohlwill, 1966,

p. 15). Thus both planning needs and the requirement for a

two-faceted environmental definition suggest the need to

expand this research perspective and develop a framework to

define the categories of environmental stimulation.

The motivational properties of environmental stimula-

tion can be specified in terms of behavior instigated by

and directed at particular environmental characteristics

(Wohlwill, 1970).

The proximal impact of environmental stimuli can be

seen in their role as sources of affect and attitude and

approach and avoidance response.

As a source of affect and attitude the environment is

functionally related to buyer predispositions. There is

growing support for the thesis that physical settings are

closely related to attitudinal and value systems, expectan-

cies, general affect, exploratory drive, and curiosity (Parr,

1966, p. 42; Fatouros, 1968, p. 520; Proshansky, Ittelson,

and Rivlin, 1970c, p. 28). In addition, urban environments

have been alternatively characterized as scenes of sensory

deprivation and overstimulation and stress (Wolpert, 1966).

The distinction is especially important when considered in

terms of the attitudinal and affective consequences of stimu-

lus variation. Causative direction has been difficult to

establish in matters of urban stimulation. Opposing views

exist with respect to whether an urban customer is alert

because he is experiencing less strain from environmental

stimuli or whether he experiences less strain because he is

alert (D. H. K. Lee, 1966, p. 85). In any case a lower

degree of strain and alertness are somehow linked and tend

to influence a broad range of intellectual activities,

including choice behavior. A definite answer in this

monotony-understimulation, complexity-overstimulation debate

would require the isolation of the relevant stimulus input

variables. Certainly the intricacies of environmental

stimulation can involve the coexistence of monotony and

complexity. Thus the physical design of a shopping area

may lack stimulus value or stimulus variation while the

sights and sounds of human and vehicular movement may, as

intense and complex stimuli, be sources of stress (Kates

and Wohlwill, 1966, p. 16; Rapoport and Kantor, 1967).

Still, principles derived from experimental research on the

motivating effects of stimulation suggest the existence of

an optimal level of stimulation (Wohlwill, 1966), defined

in terms of customer behavior with respect to that shopping


While the magnitude and directional influence of

stimulus variables remain matters for future research,

general functional linkages are being established. There

are indications, for example, that changes in urban environ-

ment produce significant modifications in individual value

systems and attitudes with respect to object, particularly

goods, orientations as well as the nature of prepurchase

decision processes (Michelson, 1970, p. 141).

The concept of optimal levels of environmental stimu-

lation suggests environmental attributes play a role as

sources approach and avoidance response. At the very mini-

mum this provides a basis for judging avoidance behavior

as the joint product of monotony and overcomplexity (Kates

and Wohlwill, 1966, p. 16). In more general terms, external

search propensities and customer spatial extensivity can be

seen to be related to the design of retail facilities and

the paths leading to them and not merely to the character

of merchandise offerings. These parameters of environ-

mental behavior are typically assumed as being identical by

gravity models (Huff, 1962a, p. 65) and are totally ignored

in goods typologies. Concern with these matters in trade

area and location analysis (Applebaum and Cohen, 1961) and

store layout (Gist, 1968, pp. 231-52; for two unusual

approaches see Krugman, 1960, and Winkel and Sasanoff, 1970)

literature is usually limited to isolated observations and

lacking in theoretical underpinnings.

Actually, customer spatial patterns are related to the

total relevant environmental situation, with goods and ser-

vice mixes acting as only one of a set of influential vari-

ables. Thus the number and distribution of retail facilities

is related not merely to the character of customer spatial

movements (Yuill, 1967, p. 107) but perhaps also to their

speed and direction (Parr, 1966, p. 143). In this manner

the correlates of multipurpose trips and the observation of

accelerated movement in the direction of a diminishing

visual inventory can be given greater significance than that

of mere isolated relationships by placing them within the

confines of a customer environmental behavior model and

studying them as approach-avoidance phenomena.

Approach and avoidance may not be the direct behavioral

counterparts of summed positive and negative valences asso-

ciated with a set of alternative environmental experiences.

A mechanism may operate to allow the customer to cope with

environmental disamenities and a disamenity threshold, which

must be exceeded before the stressful consequences of nega-

tive environmental features escalate into strain, may tend

to filter the results of environmental experience. Thus a

moderate degree of stress may induce search behavior just

as uncertainty without anxiety may be associated with pro-

ductive work (Wasson, Sturdivant and McConaughy, 1968,

pp. 35-6; Wolpert, 1966, pp. 94-5).

Affect and attitude and approach and avoidance tenden-

cies can manifest themselves in the present states and overt

behavior without immediately,or ever, percipitating any

significant behavioral consequences (Parr, 1966, p. 42).

Of course, it is possible that exposure to more variable

environments results in customers who are more perceptually

alert and behaviorally active, that urban monotony produces

the effects of long-term sensory deprivation, that percep-

tion of the urban landscape is primarily a function of

extended exposure to habituated environmental constants,


These and other distal responses to environmental

stimulation are the result of processes of adaptation or

adjustment. As a seeker and neutralizer of environmental

stimulation, the customer employs various physiological

and psychological adaption processes (D. H. K. Lee, 1966,

p. 87), which, for example, may help neutralize and surmount

the barrier properties of a path to a shopping alternative.

Incomplete neutralization, adjusted for acceptable devia-

tion from an adaption level (Wohlwill, 1968-1970), may

escalate into strain (Wolpert, 1966), yield an avoidance

response, and, over time, produce permanent modifications

in spatial choice processes and habits (Gist, 1968, p. 169).

Alternatively, a form of dynamic adjustment through

technology is possible. Thus the customer traveling in

a quiet, air conditioned vehicle has not adapted to the

sounds, odor, heat and humidity. Instead technological

adjustment has reduced the stress on the customer's adap-

tion processes. On the other hand, adaptation on the part

of a customer who must face such urban elements without a

technological shield may leave a detectable behavioral resi-

due, possibly in the form of altered arousal thresholds,

frustration tolerance, physical discrimination abilities

(Wohlwill, 1966, p. 36), etc., which characterize behavior

across a constellation of roles, including those associated

with customer behavior.

Similar adaptation and adjustment processes can result

in a group of customers becoming, to a large extent, com-

monly conditioned to environmental factors. Sharing certain

expectancies, these so-called natives accept and even posi-

tively value environmental features that may be regarded as

otherwise by non-natives, i.e. the new or visiting customers

and those, who despite their presence in the same physical

environment, share a different set of environmental experi-

ences and expectations (Kates and Wohlwill, 1966, p. 17;

Sonnenfeld, 1966). In marketing literature non-native

learning processes have been considered chiefly in terms

of brand behavior (Atkin, 19661 Halbert, 19661 Andreasen

and Durkson, 1968).

Yet, in spite of the proclivity of marketing theorists

to neglect their existence, it would seem that the stimulus

properties of the urban environment function prosthetically

to support behavior goals, through the maintenance of behav-

iorally correlated physiological states,and elicit stimulus

influence over certain behavioral topographies (Michelson,

1970, p. 26; Studer, 1970, p. 58). Environmental stimulation

enters discussions of retail theory in the form of consider-

ation of the proximate effects of store design (Krugman,

1960; Engel, Kollat, and Blackwell, 1968, pp. 474-99; Gist,

1968, pp. 231-50; Winkel and Sasanoff, 1970). There is an

absence of speculation concerning the long-term behavioral

effects of changes in urban retail structure or internal

features of retail facilities. These are vital considera-

tions in evaluating alternative urban plans and in predict-

ing the consequences of urban growth and change.

Ironically, the most systematic treatment of the stimu-

lus effects of the retail element and its urban setting are

to be found in gravity and trade area approaches. Even here,

however, such relationships are implied rather than explic-

itly treated. This is understandable insofar as retail

gravity is an empirical concept, devoid of behavioral founda-

tions and internal theoretical cohesion. Gravity statements,

after all, are neither laws nor theories but rather his-

torical accidents produced by methods designed for short-

cut approximation of the direction and magnitude of travel

movement (Isard, 1960, p. 566; Huff, 1962a, p. 65; Gold-

stucker, 1965; Moore and Mayer, 1966; D. L. Thompson, 1966,

pp. 5-6; Scott, 1970, p. 180). It is also understandable

insofar as trade area analysis, hampered by a preoccupation

with unexplained aggregated statistical regularities and

crude boundary estimation, is remarkably free of theoreti-

cal considerations. Trade area analysis, after all, has

thrived on a steady diet of descriptive studies of site

accessibility, traffic flow, extent of patronage, composi-

tion and distribution of population, income and economic

stability, and existing and potential competition and

erroneously transformed retail demand, profit and location

into sequential considerations (D. L. Thompson, 1966, pp. 6,

11-12; P. T. Nelson, 1970).

Concentrating on the effects, rather than the nature

of environmental stimulation, gravity and trade area stu-

dents have shared in a search for spatial equilibrium solu-

tions. Such efforts have influenced the development of

retail analysis by presenting the possibility of a unique

objectively derived solution acting as a terminating point

for some undefined iterative customer process. Such a

solution has meaning in marketing and urban design in that

it contributes to the construction of meaningful theoretical

and physical spatial systems (Golledge, 1970, p. 417).


The notion of equilibrium involves a balance of forces

producing a stable state in a system. General behavioral

equilibrium involves homeostasis and, within limits, repe-

titive and invariant behavior. Thus there is no net change

in behavior between successive periods of observation.

Environmental or spatial equilibrium is merely a special

case of general behavioral equilibrium. Like its counter-

parts in other behavioral subsystems, environmental or spa-

tial equilibrium can only be subjectively defined in terms

of arbitrary time slices and with respect to a braoder

behavioral model (Golledge, 1970, p. 417).

This is evident if customer spatial equilibrium is

derived from the corresponding general equilibrium states

inherent in Kotler's (Kotler, 1965) alternative models of

consumer behavior. Thus the spatial equilibrium correspond-

ing to the Marshallian model would be represented by an

arrangement of discrete market areas while the Pavlovian

variety would reflect stimulus prone, habitual spatial

tendencies of a stereotyped, rigid and repetitive nature

(Golledge and Brown, 1967; Golledge, 1970, p. 418), A

socially determined customer spatial equilibrium can be

derived from the Veblenian model but it lacks an a prior

spatial character. Only in the case of the Freudian model,

which involves only short-term regularities in spatial

behavior, is it impossible to define behavioral, and thus

to derive any spatial, equilibrium (Golledge, 1970, p. 418).

In a somewhat less systematic fashion, spatial equil-

ibrium states can also be derived from assumed customer

characteristics. In this fashion, cognitive dissonance

(Festinger, 1957) can be used to develop an equilibrium

patronage pattern which would minimize the dissonant feel-

ing following customer environmental response, in a manner

similar to the minimum regret of game theory. A presumed

characteristic of spatial rationality on the other hand,

would provide for choice of least effort or minimum aggre-

gate distance paths. The customer would move down a heir-

archy of route preferences as barriers presented themselves.

Customer environmental response would involve a variety of

transport modes but invariable rational path selection

because patronized retail elements would be those seen as

nearest or involving least effort. Equilibrium would be

disturbed only as additional information prompted attempts

to eliminate spatially irrational acts. Equilibrium itself

might involve least effort principles, single choice syn-

dromes, or habitual multiple patronage (Golledge, 1970,

pp. 418-9).

Obviously equilibrium states based on a single cus-

tomer characteristic are analytically deficient insofar as

they lack the foundation provided by the explicit morphology

and behavioral mechanisms of a general model of consumer

behavior. Thus they are theoretically barren, except in

terms of comparative statics, and essentially tautological.

They share these characteristics with most mathematical

models of customer spatial equilibrium.

Such models fall into place-loyalty, market share,

learning and stochastic perceptual categories.

In place-loyalty models consumers, assumed to be scat-

tered in space, develop habitual preferences resulting in

their patronizing one retail place to the exclusion of all

others. Generally economic and spatial rationality charac-

terize behavior and least effort tendencies operate, with

effort measured by increments to the base price of goods

resulting from movement. This form of spatial equilibrium

output can also be derived from central place, gravity, and

retail location models (Carrothers, 1956; Huff, 1964; B. J. L.

Berry, 1968).

The deterministic framework, the permanent allocation

of customers to discrete retail nodes and the finite bounda-

ries to areas of nodal dominance typically found in such

models directly conflict with empirical evidence of customer

behavior (Rushton, Golledge, and Brown, 1965; Golledge and

Brown, 1967, p. 116; Enis and Paul, 1970). In single vector

and Markov variations of market share models (Golledge, 1970,

pp. 419-20), however, explicit recognition is given to pos-

sibility of customer uncertainty and response change occur-

ing in an equilibrium situation. As a result, probability

distributions are used to specify patronage proportions over

retail alternatives. Frequency of patronage can be variously

established by studying the frequency of individual patronage,

averaging the proportion of times which group members

patronize retail places, or employing store characteris-

tics as surrogate indicators. In simple models the equilib-

rium output is represented by a constant probability vec-

tor. Huff's (Huff, 1962b, 1963, 1964, 1966) construct is

one example of single-vector market share models.

Although such models seem consistent with the probabilis-

tic nature of customer environmental behavior, they do not

specify how the market-sharing principle was established.

Since vector elements are usually based on cross-sectional

data, we have the paradox of an equilibrium solution pro-

vided without evidence that consumers are in equilibrium.

Markov variations help reduce the first of these objec-

tions by assuming patronage is based on habit rather than

conscious preference behavior. Initial behavior is thought

to be modified by the rewards associated with trial and

error, i.e., the strengthening of rewarded responses and the

extinction of non-rewarded efforts. While this series of

events might be represented by a single-vector model subject

to changes through time, the more general form is a first-

order Markov model which describes the equilibrium state by

stable-state proportions. Customer switching and staying

tendencies are summarized in a matrix of transition proba-

bilities. Thus the equilibrium state evolves from an ini-

tial patronage configuration. The operation of this model,

by emphasizing the effect of last patronage on next patron-

age probability, captures certain facets of the role of

learning and experience in producing spatial equilibrium.


The relationship between learning and spatial equilib-

rium is given explicit consideration in linear operator

learning models (Golledge, 1970, pp. 420-2). One-element

models treat reward and nonreward as having an equal effect

on patronage. Using a first-order difference equation, this

model depicts a customer's store choice as being equal to

the probability of that choice in the last period plus an

increment proportion to the maximum possible increase in

patronage probability. Such models employ empirical or

Bayesean estimates of last period patronage, an assumption

that patronage is a function of the number of reinforcing

experiences, and a learning parameter that determines the

rate of approach to an asymptotic equilibrium state.

The more comprehensive two-element models treat reward

and nonreward as having different effects on patronage and

use two learning parameters to describe the rate at which

patronage probability is incremented or diminished by store

choice or rejection. On the assumption that learning is

always incomplete, these models provide for upper and lower

asymptotic patronage limits.

The problem of numerous retail alternatives and the

distribution of decreased patronage can be handled by using

vectors specifying asymptotic patronage limits for each

feasible place in an n-operator model. Unfortunately, the

myopic concern with past choice prevents a consideration of

the total marketing efforts of different retailers and their

effects on patronage. The most important defect is that


patronage, regardless of its consequences, always increases

the selection habit and only nonpatronage tends to limit

habit formation (Kotler, 1968, pp. 281-2; Golledge, 1970,

pp. 421-2).

There remain, however, fundamental questions as to

the validity of the concept of patronage equilibrium. It

is obvious that exclusive patronage is hopelessly inadequate

as a description of customer spatial equilibrium. Yet it

may also be that observed customer microgeographic varia-

bility invalidates concepts of individual patronage equilib-

rium tendencies as well as spatially competitive equilibrium

between retailers (Mason and Moore, 1970-1, pp. 36-7). One

need not adopt this extreme position to recognize that cus-

tomer spatial equilibrium patterns are not immediately

determinate. Certainly search and shopping tendencies form

the basis for a customer's eventual system of optimal spa-

tial connections. The duration of pre-equilibrium activi-

ties is a function of goods and store type and frequency of

purchase and visit. It may be that in some cases this period

of non-optimal spatial connection may never end (Berry and

Garrison, 1958b, pp. 119-20) while in other instances the

dynamics of retail competition and urban change result in

short-lived equilibrium states.

There is clearly a need in marketing for theory con-

cerning non-equilibrium states and the associated perceptual

processes which influence customer movement. At present the

temporary character of urban patronage equilibrium is best

captured by stochastic perceptual models (Golledge, 1970,

pp. 422-3). These models employ the product of a response

vector, containing the relative ratings of salient retail

attributes, and a rating scale matrix, containing store rat-

ings along salient attribute dimensions, as an index of

patronage probabilities. Although the use of normalized

average perceptions limits these models to aggregate appli-

cation, such a scale problem is not insurmountable. What

is crucial is that these models explicitly handle the influ-

ence of temporal variations in marketing mix inputs and pro-

vide for only temporary equilibrium states. The integration

of the insights provided by stochastic perceptual models

into a framework which comprehends the contextual and stimu-

lus properties of the urban environment would have several

advantages. First it would allow marketing theory to focus

on nonequilibrium customer processes and behavior. This

provides a basis for the analysis of nonequilibrium spatial

systems in urban planning. Second, this integration would

relieve customer behavior constructs of their deterministic

character by transforming patronage equilibrium from an

assumption and a matter of eventual fact into one of a num-

ber of testable propositions. Third, it would free retail

theory of the shackles of environmental neutralism fashioned

in an era concerned with rural, not urban commerce. Finally,

it would be theoretically parsimonious in that it allows us

to avoid the useless complexity required in models based on

behavioral, spatial, and physical determinism.

A Four-Element Intersystem Congruence Model

The four previous requirements for a model of envir-

onmental behavior suggest the existence of four essential

elements. There is an individual decision maker (the cus-

tomer), whose environmental response behavior (path and store

choice) is a function of an environmental situation (which

includes, but is not limited to, store environment) and an

environmental decision process (Blaut, 1969, pp. 1-9; R. M.

Downs, 1970).

Reciprocity (between the customer and the environmental

situation) provides the environmental decision process with

a dynamic and evolving character. The dual definition of

the environmental situation suggests that environmental per-

ception is a major parameter in the customer-environment

interchange and thus in the environmental decision process.

Together with the contextual and stimulus properties of the

urban matrix, environmental perception forms the behavioral

basis for path and store choice, i.e. customer environmental

response. This suggests the existence of three sets of

feedback loops. Primary feedback involves the consequences

of continual interaction with, and redefinition of, the urban

environment. Secondary feedback involves the stored conse-

quences of current environmental response. Tertiary feed-

back involves residual effects of cumulative environmental

responses and experiences.

In this context it becomes apparent that the descrip-

tive heuristics of gravity-based projections of customer

spatial behavior form an inadequate basis for a theory of

customer environmental behavior. It is also plain that

attempts to show up such theoretical deficiencies, by employ-

ing assumptions of customer behavioral tendencies derived

from an economic man concept, have amounted to a normative

prescriptive explanation of customer environmental behavior

leading to a mechanistic pairing of customers with retail

facilities (R. M. Downs, 1970, p. 14).

A distinguishing characteristic of a customer environ-

mental behavior model is the explicit recognition of the

fact that the bonds between urban activities are so pronounced

and exert such an integrating effect that retail and customer

spatial analysis must treat these activities as bound together

in an urban complex, rather than separately (Isard, 1960,

pp. 673-4; A. Mayer, 1967; Schnore and Lampard, 1967, p. 24).

Thus the model's four criteria as well as its essential ele-

ments present the consumer as relating and reacting to the

structure of the psychologically relevant urban environment

rather than to isolated stores or other individual physical


An examination of the character of this relationship

provides the theoretical frame necessary to conceptually

link the model's elements. Earlier consideration of impli-

cations of the model's first four criteria suggested the

rejection of both environmental (spatial) determinism and

dominance by non-physical (cultural, social, and psychologi-

cal) systems. A long tradition of non-physical system

dominance in social science theory and research has served

to indicate that this view is barren of environmental con-

sideration and, for our purposes, theoretically sterile.

Environmental determinism would suggest that customer

environmental response is determined by the arrangement of

urban physical space. Research dealing with other role

types and behavior subsystems lends initial support, by

inference, to this contention. Closer inspection, however,

reveals that perceived, induced, or actual homogeneity of

residents is often responsible for common response patterns.

Even these similarities tend to diminish as time passes

(Heyman, 1964; Michelson, 1970, pp. 168-90), indicating that

learning and adjustment processes modify apparent determin-

ism. In these cases as well as those involving customer

behavior, environmental determinism tends to factor out the

significant processes of environmental decision making. Thus

it is necessary to distinguish between determinism and influ-


Certainly, an analysis of the customer-environment

relationship should recognize that certain "states of vari-

ables in one system can coexist better with states of vari-
ables in another system than with alternative states"

(Michelson, 1970, pp. 25-6). When examples of such congruence

The congruence concept suggests that different psycho-
social characteristics of urban residents are consistent
with different urban physical systems, i.e., they tend to
coexist better with certain physical arrangements than with
other variations of the urban matrix.

are found with respect to customer environmental behavior

and analyzed in terms of the previously established criteria

and elements, we have developed the basis for an intersys-

tems congruence model (Michelson, 1970, pp. 25-30). This

model would suggest the broad limits within which customer

cultural, social and psychological characteristics and urban

environmental features systematically articulate with one


Congruence is consistent with requirements for envir-

onmental definition because it allows customer variability

and choice to operate within the limits of differential per-

ception without ruling out the basic relationship, between

the urban physical system and customer non-physical systems,

imposed by the environment. Thus it is possible to observe

how particular retail structures within urban environments

articulate with customer values, attitudes, and spatial ten-

dencies. Certainly any mismatch or incongruence may be

properly maintained despite its dysfunctional nature because

intermediate variables intervene or because it is the result

of more functional relations (Alexander, 1970; Haythorn,

1970). The concept is used to convey the idea that poten-

tial customer spatial diversity is constrained by the limi-

tation of not only common cultural, social, or psychological

characteristics but by existing or future urban form (Carr,

1970, p. 521; Michelson, 1970, pp. 24-6; Studer, 1970).

An intersystem congruence model will facilitate thp

organization of relevant variables in designed urban


environments in accordance with specified behavioral requi-

sites. The range of such alternative spatial outcomes has

been generally disregarded in planning research. The per-

tinent question now is not the relative importance of designed

retail or urban environments but rather whether different

plans for an urban area and its retail structure result in

corresponding differences in customer behavior (Gutman, 1966,

p. 113; Stagner, 1970, p. 197).

The design of the retail structure or any other urban

physical subsystem cannot, however, be automatically linked

to time- and space-bound precepts of future users. A dis-

tinction must be drawn between mental and experiential con-

gruence. Mental congruence is essentially expectational and

involves beliefs concerning the ability of a spatial pat-

tern to accommodate a characteristic mode of environmental

behavior (Michelson, 1970, pp. 30-1, 204-6). Studies of

mental congruence seek to discover, for example, what cus-

tomers want in terms of the spatial character of retail

facilities and why these features seem desirable. Thus they

are limited to uncovering regularities in expectations regard-

ing consequences arising from the creation or maintenance of

a particular physical environment. Results are further

qualified because responses are a function of existing

environmental alternatives and limited customer environmental

experience (Carr, 1970, p. 521). Experiential congruence,

while also related to expectations (Sonnenfeld, 1966; Wol-

pert, 1966; Michelson, 1970, p. 30), involves actual behavior

and how well an environment accomodates actual character-

istics and behavior of users (Michelson, 1970, pp. 31-3,

206-13). Customers may not be aware of the existence of

experimental congruence, but their actual experiences can

indicate the constraining or enabling influences of their

urban and retail environment.

The congruence concept is the product of a realistic

interpretation of the inherent limitations of urban envir-

onmental research. The concept is fashioned to cope with

otherwise inconclusive results of studies which find it

impossible at any single moment to measure the impact of the

total environment (Dyckman, 1961). Design, by its nature,

involves the simultaneous and sequential combination of many

elements to form a total urban system. The task is compli-

cated by the probable nonlinear, discontinuous relationship

among variables in the urban system and between these vari-

ables and the values of any objective function. Given the

possibility of developing new design elements and the exist-

ence of relatively few external constraints, the solution

space for the design of an urban system has undefined dimen-

sions which prevent the total analytic treatment of the

planning process to reach a true optimum result (Harris,

1967). The best hope is for suboptimal urban environments

in which desirable adaptive behavior is made possible by

intersystem congruence.

Given even linear functions, existing technology, and

a suitable number of constraints, the specification of

optimal urban environmental conditions remains impossible

(Proshansky, Ittelson, and Rivlin, 1970b, p. 278, 1970c,

p. 494). The current level of theoretical development and

empirical ignorance does not allow us to deal with conflict-

ing behavioral goals or even specify the role and functions

of an urban retail structure (Duhl, 1963, pp. 139-40; Gut-

man, 1966, p. 111). Thus a retail or total servo-environment

composed of behavior-contingent subsystems will remain an

ideal (Sommer, 1969, pp. 145-54; Alexander, 1970; Studer,


It is possible to distinguish between environmental

characteristics which facilitate and obstruct various behav-

ioral goals. The congruence concept provides a suitable

vehicle for such analysis while acting as a safeguard against

the adoption of restricted causal chains or functional rela-

tionships which link particular physical stimuli to particu-

lar environmental responses (Kates and Wohlwill, 1966).

The limitation created by focusing on the retail ele-

ments in the urban system and on the customer role type may

appear artificial. Yet a holistic vision of the urban envir-

onment should not produce futile attempts at holistic manage-

ment. The feasible approach is through general systems

theory which, like ad hoc methods, deals with elements and

subsystems, but unlike the ad hoc approach deals with them,

as much as possible, in relation to the total system of

which they are interrelating components (Caldwell, 1970,

P. 75).

This approach provides the basis for intelligent

environmental analysis and choice. It does not provide

a means to design self-maintaining urban retail environ-

ments, Recent planning experience has made it evident that

environmental intervention increases, rather than diminishes,

the need for environmental management (Willmott, 1967, p. 3961

Willis, 1969, p. 189). Administrative necessity is under-

scored by the dynamic properties of congruence, the subop-

timal nature of urban design, and the need to provide for

future accommodations. The potential insights provided by

an intersystem congruence model have a value content which

is independent of political and economic context (Applebaum,

1965). Even unlikely retreat into policy neutralism or

avowed pluralism, two complementary anti-planning positions

(Erber, 1970, p. xv), will not diminish the value of increased

understanding of customer environmental behavior.



The previous analysis suggests that there are factors

outside the urban retail spatial system, i.e., the number

and type of retail facilities and their spatial distribu-

tion, which influence customer environmental behavior. To

understand how customer environmental behavior is related

to the urban retail spatial system and to explore the con-

sequences of differential spatial organization, we must

avoid the assumption that the identity and relatedness of

retail elements can be derived from a total concentration

on urban arterial networks. Instead we must acknowledge

that the space within which urban residents carry out their

daily activities is not merely defined by the physical bar-

riers which restrict motion and/or the reception of urban

stimuli. Urban space is also defined by the behavior of

individuals who occupy that space (Lynch and Rivkin, 1959;

Duhl, 1963; Stea, 1965). Conventional investigations of

the man-environment interchange have paid insufficient

attention to the human side of the equation. A review of

past research reveals that models based on vaguely defined

assumptions of rationality or focusing on objective pushes,

pulls, and opportunities fail to explain a considerable

portion of the variation in environmental response.

As Stea and Downs (Stea and Downs, 1970, p. 5) have

noted, it is this unexplained behavioral variation, which

was once attributed to so-called irrational forces and con-

sidered to be the noise of earlier research, which now

becomes the signal or subject matter in the current study

of environmental behavior. The basis for this research

knowledge lies in an understanding of how urban residents

(in our case, customers) comprehend the urban environment,

i.e. interpret the physical and nonphysical nature of the

surrounding spatial field. The transactions between the

customer and this field (as defined by the reciprocity con-

cept) are a function of configurations of stimulus and con-

textual properties of the urban environment (as defined by

the dual definition concept) as they are distributed in

space (Beck, 1970, pp. 134-5).

It is dangerous to assume that the space within which

customers move has entirely objective geometric properties.

The distinct difference in meaning between various urban

areas (T. Lee, 1968; Carr, 1970; Strauss, 1970) and direc-

tions (Kepes, 1961; T. Lee, 1968, 1970; Lowrey, 1970) give

urban space a nonhomogeneous character (Kepes, 1961; latridis,

1966; DeJonge, 1967-8, p. 10; Horton and Reynolds, 1969).

These differences are compounded by significant variations

in the manner in which different urban groups comprehend and

use their environmental surroundings (Fried and Gleicher,

1961; Duhl, 1963; Webber, 1964; Beck, 1970; Strauss, 1970).

Unfortunately, the observation of such variation has not

been accompanied by the identification of those variables

which link environmental comprehension to behavior. This

restricts our ability to describe the nature of the envir-

onmental decision process which underlies customer environ-

mental behavior.

Thus in present exploratory research efforts it is

both necessary and desirable to focus on the subjective

basis of the environmental decision process, urban environ-

mental perception, and, in Carr's (Carr, 1970) terminology,

to attempt to understand the "city of the mind." Certainly

the ability to eventually develop a model capable of cap-

turing the results of customer environmental choice depends

on an understanding of how the objective urban landscape is

subjectively interpreted by the customer (D. L. Thompson,

1966, p. 9).


The purpose of the present research effort is to develop

an exploratory model in order to define, study, and measure

customer environmental maps, i.e. the conceptual or imaged

space which represents a customer's understanding of his

urban retail environment. In order to develop the frame-

work for such a tentative model, it is necessary to review

some of the attempts to investigate the mental maps of

Impressions of large physical areas and urban areas
have variously been referred to as mental maps, cognitive
maps, schemata, conceptual spaces, and imaged spaces.

urban residents so that we may develop assumptions regard-

ing the special nature of customer environmental perception

and outline some of the principles which customers use in

ordering retail conceptual space. We will then discuss the

nature of customer environmental perception and some of the

functions of customer environmental maps and contrast cer-

tain key notions with earlier psychological research. Using

customer environmental maps as a theoretical construct and

experimental output mode, we shall discuss the nature of

relevant metrics and develop a summary statement and defini-

tion of customer environmental maps. It will then be pos-

sible to develop certain testable hypotheses. Hopefully

this effort will contribute to our understanding of the cus-

tomer environmental decision process and thus to the intel-

ligent planning of urban retail facilities.

Related Research

The current interest in urban imagery and maps as per-

ceptual phenomena can be traced to the pioneering research

of Lynch (Lynch, 1960), who employed the concept of an image

as a cognitive representation of a large urban physical

area. Although the image concept had earlier applications,

its use in an urban setting was both new and significant,

especially since, as shown below, it involved an important

departure from the limited setting of conventional psycho-

logical studies of the perception of stimuli. In studying

the comprehension of the city-scape by residents of Boston,

Los Angeles, and Jersey City, Lynch, as a planner, was

primarily interested in uncovering the relationship between

respondents' urban images and the actual physical form of

the city. Comparing consensual urban images to visual

reality, Lynch sought to offer principles for urban design.

Using respondent's map sketches and verbal responses, Lynch

constructed the urban images or maps which residents seemed

to employ in their daily decision processes. These maps

revealed important differences in the manner in which resi-

dents organized city dimensions and images and significant

alterations and deletions of urban landmarks, districts, and

travel routes. Subsequent applications of Lynch's methodol-

ogy in other urban settings (DeJonge, 1962; Gulick, 1963;

Appleyard, 1969, 1970; Porteus, 1971) have substantiated his

major conclusions and extended knowledge concerning the com-

prehension of an urban matrix in terms of both image for-

mation and image structure. As Thompson (D. L. Thompson,

1966) has observed, Lynch's seminal efforts have important

implications for marketing theorists because an insight into

the subjective reactions of residents to urban environmental

complexity may carry important clues for analyzing shopping

behavior and the spatial distribution of patronage.

Another development in the study of urban imagery is

the investigation of how cognitive representations of the

urban environment are formed. In this connection a concern

with what people look at as they move through the city, how

they remember such contacts, and how they use their memory

in reconstructing that experience (as they might in carry-

ing out daily activities) has spurred an interest in the

cognitive mapping processes associated with the experience

of moving through the city on an expressway as a driver,

occupant, or commuter. The findings of Appleyard, Lynch

and Myer (Appleyard, Lynch and Myer, 1964) and Carr and

Schissler (Carr and Schissler, 1969) in this area have

helped uncover the primary perceptions of an urban highway

sequence, the sensations of space and motion, and the most

readily identifiable elements of attention in such situa-

tions. Although primary interest has centered on the psycho-

logical effects of urban highway design, this literature

carries important implications for marketing analysts inter-

ested in the perception and memory representation of such

trips as they relate to expectations and patterns of look-

ing. These studies have focused on the roles of such vari-

ables in organizing and transforming sequential trip per-

ceptions into cognitive environmental representations.

A third body of related research has arisen in connec-

tion with the developing interest in the duality of the

physical and nonphysical neighborhood. Studies eliciting

phenomenological representations of urban neighborhoods

(T. Lee, 1966, 1968; Willmott, 1967; Ladd, 1970) have

employed these outputs as primary indicators of the social

and psychological contents of neighborhoods and, indirectly,

of the limited nature of individual action spaces. Although

their methodology and objectives have differed, neighborhood

mapping studies have shared three common features. First,

they have provided indications of the potential information

value of maps as inputs in the urban planning process.

Second, they have helped to reveal sources of individual

and group variation in phenomenological representations of

urban space. Finally, they have contributed to knowledge

concerning environmental perception by providing insights

into the cognitive processes involved in the comprehension,

organization, and representation of an urban area.

The general theoretical foundation for a model contain-

ing these and other aspects of urban images or maps lies in

understanding of the psychological basis of space perception

and the learning processes involved in environmental inter-

action (Harvey, 1969a, pp. 192-7). While definite conclu-

sions are not at hand, recent concern about these basic

matters has led to empirical studies and scientific specu-

lation (Shelton, 1967; Tilly, 1967; Steinitz, 1968; Blaut,

1969; Blaut, McCleary, and Blaut, 1970; Ladd, 1970). It

would seem that mapping processes are an important component

in environmental learning. It would also appear that images

or maps of urban areas are not isolated phenomena but are

directly related to the observed ability in children to

generate, represent, and utilize cognitive maps of more

limited spatial environments.

An Exploratory Model for the Measurement of
Customer Environmental Maps

While the previous literature does not directly deal

with customer environmental behavior, it does provide clues

as to the possible variables and mechanisms involved in the

urban environmental perception of customers. Such percep-

tion is regarded here as a set of micro-geographic percep-

tual experiences which are used by customers to form a macro-

environmental strategy of behavior (Blaut, McCleary, and

Blaut, 1970, p. 337). At present the most important priority

is to develop a conceptual framework to link the variables

and mechanisms in question and to provide a guide for future

research. A useful reference in this connection is Stea's

(Stea, 1969) experimental model for studying the mental maps

and conceptual spaces associated with the perception of large

physical environments. Stea's model, the product of older

psychological research and recent findings concerning envir-

onmental perception, is intended to serve as a guide for

organizing experimental research. The model is neither urban
in orientation nor readily applicable to customer behavior.

The present objective is to modify the Stea model, in terms

of the general model of environmental behavior and insights

provided by research in psychology, sociology, behavioral

geography, and customer behavior, in order to develop a new

Stea (1965) has applied the mental map concept to
more limited contexts, suggesting that different mental map
models can be developed for different environmental settings.

construct suited to the study of the environmental percep-

tion of urban customers.

Basic Assumptions

Initially it is necessary to make the three following


First, customers form conceptions of significant por-

tions of the urban display or objective urban environment,

that are too large to be perceived or apprehended at once.

The retail component of the urban display is the total

retail store set (TRSS), which contains all retail alterna-

tives. For any expected pattern of customer behavior there

is an existing group of retail alternatives, the retail

store subset (RSS). Taken together, all HSS's are collec-

tively exhaustive, but not mutually exclusive, subsets of

the TRSS.

Second, customer conceptions of the objective urban

retail environment contain two basic categories of retail

elements. There are, first of all, those retail elements

in the objective urban landscape of which the customer is

aware, the total comprehended retail store set (TCRSS).

An expected pattern of behavior refers to a customer's
expectations regarding the character (instrumental versus
congenial) and nature (purchase versus search) of a forth-
coming trip. (See Alderson, 1965, pp. 144-51.)

The term element refers to the fact that, in a cus-
tomer's conception of urban retail reality, retail alterna-
tives are conceived as macroenvironmental information cate-
gories, i.e. cognitive categories into which information
about the retail spatial system is coded (see Stea and Downs,
1970, p. 8).

For any expected pattern of customer behavior there is a

corresponding group of potential retail elements of which

the customer is aware, the comprehended retail store sub-

set (CRSS). Since a customer may be aware of a retail

element yet not consider it to be an acceptable alterna-

tive, CRSS's, taken together, are neither collectively

exhaustive nor mutually exclusive subsets of the TCRSS.5

A second category, the total evoked store set (TESS), con-

tains the elements in the urban retail landscape which the

customer considers as alternatives for all his expected pat-

terns of behavior. For each expected pattern of behavior

there is a corresponding subset of acceptable retail ele-

ments, the evoked store subset (ESS). Taken together, a

customer's ESS's are collectively exhaustive, but not mutu-

ally exclusive, subsets of the TESS.

Third, the customer's conception of the urban retail

landscape, which contains the aforementioned retail ele-

ments, may not be spatial but, by means of certain mechan-

isms, orders retail elements, which are themselves spatial


Ordering Mechanisms

While the exact nature of these ordering mechanisms

is not known, it is possible to suggest certain principles

5Since a retail facility can operate as a landmark,
orientation reference, etc., a customer may be aware of the
physical existence of a retail facility but not its func-
tional nature as a retail unit. Such a facility would be
included in his TCRSS but not his CRSS.

which guide the ordering of elements in imaged retail

space. Because of the exploratory nature of this analy-

sis, it is best to confine our remarks to elements in the

TESS and ESS. There are five ordering principles.

First, hierarchies of retail elements are established

by the customer. Thus some alternatives, for a variety of

reasons, are considered to be more important than others.
Several hierarchical arrangements may coexist, at the same

and/or different times.

Second, the retail elements are spatial entities and

thus imaged retail space, while not necessarily continuous,

has boundaries. Since this perceived space (the customer

environmental map or CEM) ends somewhere, the extent of the

CEM tends to limit additions to, as well as reflect the

selective nature of, an ESS.

Third, the retail elements located within such space

are conceived as points, even though they may possess dimen-

sion and represent cognitive information categories. As

points, they possess certain properties which facilitate

comparison and measurement.

Fourth, because of their nature as points and macro-

environmental information categories, the ESS elements in

a CEM exist in relation to each other. This pattern of

The coexistence of several hierarchical arrangements
is the result of the customer coding retail macroenviron-
mental information along multiple salient attribute dimen-
sions and being capable of a number of expected patterns of

relationships constitute a mental geographic set (MGS)

which represent the dependent variables and conceptual

metrics of a CEM. Elements in an ESS have objective coun-

terparts in an RSS, although some ESS elements may no

longer, or never have, existed. The RSS elements are also

related to each other in terms of a similar, but objective,

set of relationships referred to as a physical geographic

set (PGS).

Fifth, the connectivity between any two points is a

function of the actual and perceived route between them as

well as barriers along that route, physical or otherwise,

including availability of a transport mode, terrain, changes

in ethnic or soclo-economic character of intervening resi-

dents, etc. The ordering mechanisms reflect the reciprocal

nature of the man-environment interchange in that 1) the

properties of the route are a partial function of the nature

of the points it connects and 2) the relationship between

points is a partial function of the barrier properties of

various routes.

The Nature of Customer Environmental Perception

Ordering mechanisms help the customer form a conceptual

representation of a relevant portion of the urban retail

environment or TRSS. In order to further understand the

nature of these ordering mechanisms it is necessary to

clarify the nature of customer environmental perception.

The notion of perception has been subject to numerous

meanings and applications in different disciplines. In

marketing perception remains a highly ambiguous concept

which has been associated with a range of subjects extend-

ing from the personal outlook of a consumer to his compre-

hension of a product, price, firm, commercial message, etc.

In each case the implied definition differs. The argument

has been made that a precise definition of perception is

neither necessary nor desirable (Harvey, 1969b). While there

may be some merit in this position, we can make a limiting

statement to the effect that perception involves some inter-

action with or transaction between the individual and the

environment, i.e. he receives information from the external

environment and somehow modifies his experience and/or

behavior. Engel, Kollat and Blackwell (Engel, Kollat, and

Blackwell, 1968, p. 79), for example, define perception as

"the process whereby stimuli are received and interpreted

by the individual and translated into a response." In a

spatial context, Harvey (Harvey, 1969b, p. 52) has suggested

that perception is "the central node in a network which

brings together cognitive processes and environmental stimuli

and which projects to the act of decision." Since it has

proven difficult to develop more specific definitions, it

would seem best to define perception, as it is used in the

present context, operationally, i.e. in terms of the CEM.

The idea that customers perceive portions of an urban

landscape tends to invest the concept of customer environmental

perception with certain properties. Cities, being large

physical areas, cannot be perceived in the usual sense. A

customer can only appreciate urban space by scanning it

temporally, can only act in urban space successively, and

thus can only perceive it by forming a collective unity of

successive impressions of urban form and content (in the

form of a CEM) from a temporal sequence of interactions with

the environment. Thus an important component of customer

environmental behavior is a process of environmental mapping,

in which retail macroenvironmental information is transformed

from time integration of sensory inputs into a cognitive

map (T. Lee, 1966; Blaut, 1969; Stea and Downs, 1970).

Environmental perception is therefore unlike other forms of

perception in that the basic unit of vision is a dynamic

sequence of vistas rather than a fixed spatial location.

The Functions of a CEM

A customer's repeated transactions with the urban

environment provide him an almost continuous input of physi-

cal and nonphysical information. Due to the previously

discussed selective tendencies, these inputs are neither

allotted equal storage space nor stored indefinitely. They

do however have the potential capacity to modify the image

or environmental map which the customer has built up over

time. In its role as a selector and modifier of relevant

environmental inputs, a CEM acts as a substitute for direct

perception of the urban retail environment. In this surrogate

capacity, the environmental map transforms complex macro-

environmental information, derived from temporally succes-

sive customer experiences, into an analogical form which

the customer can more easily assimilate and manipulate

(Blaut, 1969, pp. 14-18; Blaut, McCleary, and Blaut, 1970,

p. 338). This process conjoins three types of environmental

information concerning any relevant retail element, cogni-

tive meaning, distance, and direction. This is accomplished

by coding or mapping this triad of information in the form

of a CEM. The resultant set of relations among ESS elements

is the MGS. Eventual customer environmental behavior is a

function of an opposite process of decoding or unmapping.

In accomplishing the transformations involved in coding and

decoding the CEM performs three essential functions. In its

reduction function it builds a personally relevant model

from fragmentary real world perceptions. In its rotation

function it translates the imagined eyeview to an overhead

projection (hence the reference to the ordering of relevant

elements or cognitive categories as points). Lastly, it

simplifies urban environmental complexity and thus performs

an abstraction function.

The CEM as a Theoretical Construct and Experimental
Behavioral Output Mode

The supposition that an individual somehow perceives

a large area by forming, in the previously described manner,

an image map of that area has historical roots which reach

back over fifty years. Trowbridge (Trowbridge, 1913), for

example, presented a link between perception and imagery

in an early investigation of human orientation and so-called

imaginary maps. He suggested that two fundamental methods

of orientation guide spatial behavior. Domi-centric orien-

tation, utilizing the home origin as a chief reference in

guiding spatial behavior, was considered to be dominant in

the home vicinity. In other cases ego-centric orientation,

which involves learning to use objects or points on the

horizon corresponding to compass point directions was thought

to be employed. Trowbridge suggested that erroneous impres-

sions of compass point direction formed the basis for an

incorrect ego-centric orientation and thus for an imaginary,

or erroneous, map of the environment. Shortly thereafter,

in a separate but related set of investigations, neurologi-

cal researchers found evidence that individuals develop a

personal spatial organization, in the form of a "body schema"

or "body image" (T. Lee, 1966, pp. 22-4), which seems to

exist in partial independence of physical modification.

Both cases suggest a tendency to form personal images

of a physical entities which are often at variance with

reality. These ideas have greater significance when inte-

grated with Tolman's (Tolman, 1948) notions of mapping and

decision behavior. Tolman employed the term cognitive map

to refer to a nonliteral, but conceptually spatial, mapping

process, which he observed in rats as they connected paths

and goals. He also extended the concept to human behavior

as well. The customer environmental map differs from

Tolman's cognitive map in a number of respects but retains

an original Tolmanian characteristic in that it represents

cognitive activities of unknown nature but known effect and


Thus the CEM is presented here as a theoretical con-

struct of potential research value to marketing analysts

and planners. It is a construct which they may use rather

than believe in. The term "map" refers to a nonliteral, but

conceptually spatial, analogy. The analogies involved are

matters of process and function rather than product and

structure. The customer is seen as unmapping or decoding

stored environmental information in answer, for example, to

questions regarding the cognitive meaning of a retail ele-

ment, the location of an element in the ESS, or what element

in the TRSS is located at some point. Naturally the actual

manner in which the customer codes or engages in mapping to

transform information into storageable form is hidden from

observation. This is why the process analogy, i.e.,the

function performed by the CEM, has been stressed. Besides

this role as a theoretical construct, the map concept is

also employed here as an experimental behavioral output for

research. In this latter capacity, as Blaut (Blaut, 1969)

has noted, the map is a material model of unobservable cog-

nitive processes or "the unobservable black box into which

mapping operations are presumably plugged in like lamp


As previously noted the use of the term perception

in relation to environmental maps involves a different mean-

ing from that used in many psychological studies. Such

studies have focused on eidetic imagery, i.e., "absolute pre-

cision of memory for things once seen, a visual definition

of a visual presentation of a visual stimulus" (Stea and

Downs, 1970, p. 3). The surrogate nature of environmental

perception is quite unlike that associated with eidetic

imagery. In addition, the CEM, like Tolmanian maps, is built

from more than visual inputs. As the psychologist Welsh has

noted (Welsh, 1966, p. 423), perception of the urban envir-

onment involves multiple sense modalities. Abse (Abse, 1966,

pp. 423-4), a psychiatrist, suggests that organization and

meaning is given to these various sensations in a process

of environmental perception involving associatism and Gestalt

similarity.7 Together with the-time integrated surrogate

nature of the CEM, the fact of multiple sense inputs supports

the position that the CEM's images are not entirely eidetic.

Since spatial behavior is not limited by visual experi-

ence and since urban images do not appear to be entirely

eidetic, it seems best to presume that the CEM is not entirely

graphic. The choice of graphic maps produced from memory, in

the tradition of Lynch, as an output mode and one type of

overt response from a subject does not mean that such maps

7These ideas are especially interesting when viewed in
connection with DeJonge's (1962) findings that urban images
tend to be formed according to Gestalt guidelines.

are presented as a reification of a customer's CEM (Blaut,

1969, p. 10; Stea and Downs, 1970). Although he also

employed verbal responses, the appeal of Lynch's maps has

made it easy to equate cognitive representations of the

urban environment with graphic outputs. Hopefully we shall

avoid this pitfall. While there is no necessary correspond-

ence between mental and cartographic maps, the geographic

reification of the term "map" has, as previously indicated,

been coupled with the eidetic reification of "image." We

continue to use these terms for reason of consistency. Prob-

lems of interpretation will be avoided if the role of the

CEM as a theoretical construct and experimental output mode

is noted.

Urban and eidetic imagery also differ in that a source

of differential environmental perception is the existence

of perceptual opportunities. The fact that urban images are

only partial (Lynch, 1960) and that residents in different

areas often have mutually exclusive urban images (Michelson,

1970, p. 45) may be caused by the fact that urban spatial

behavior is largely confined to a limited action space.

Movement within this space as well as its dimensions, appar-

ently related to the life cycle and social class of its

inhabitants (Fried and Gleicher, 1961; Duhl, 1963; Webber,

1964; Willis, 1969; Beck, 1970; Strauss, 1970), place defi-

nite restrictions on perceptual opportunities and leads to

a highly selective image of the city and its elements (Lynch,

1960; Tilly, 1967; W. A. V. Clark, 1969). It is this

"awareness space" or what Clark (W. A. V. Clark, 1969,

p. 39) has called a mental map of the action space that

is used in making household decisions.

A lack of knowledge concerning brain storage and

neurophysiological processes prevents us from specifying

the exact actual nature of a CEM (or any mental) map or

even from directly testing for its existence. As a hypo-

thesized component of environmental behavior, customer map-

ping, like other mental processes, can only be studied

indirectly. Thus we utilize the role of the CEM as a con-

struct and output mode to study how customers appear to

code and decode relevant macroenvlronmental behavior for

purposes of spatial behavior. If customers behave or

describe their behavior in accordance with the map frame-

work, then its application as a theoretical construct is


Measurable Conceptual Metrics of the CEM

Adhering to Stea's original contention, it is suggested

that CEM's are measurable in that the actual and imagined

behaviors which operationally define such maps have discernible

metric characteristics. Empirical objectives thus focus on

measuring deviations in MGS metric estimates from correspond-

ing PGS relations and deviations of one metric estimate from


The mental geographic set represents the subjective

pattern of relationships between points in a CEM. CEM points

are related in terms of 1) absolute location, 2) relative

location, and, by implication 3) the extensity of the CEM,

and 4) interposed barriers. CEM points themselves have

been presented as ESS elements. In that capacity they are

macroenvironmental information categories which also have

measurable properties.

Absolute location

As a component of the MGS, absolute location is deter-

mined by absolute distance, a subjective phenomena only

generally corresponding to, and not necessarily equaling,

the objective absolute distance measure (For a contrasting

view see Lowry, 1970, pp. 52-3) of the PGS. The latter

absolute distance function, "d," defines the absolute loca-

tion of retail elements in urban spatial reality, U, the

actual metric space containing TRSS elements. If we con-

sider r, s and t to be three retail elements or points in

U and d(r,s) to represent the absolute distance between r

and s, we may say that (U,d) is a metric space insofar as

three conditions are satisfied:

(1) d(r,s) 0; d(r,s) = 0 if and only if r = s

(2) d(r,s) = d(s,r), i.e. distance is symmetric

(3) d(r,s) + d(s,t) d(r,t), the triangular ine-


An important question is whether the imaged retail

space of a CEM is metric in the above sense. To form ten-

tative conclusions, we must first understand the nature of

the absolute distance concept.

In its simplest form the absolute distance measure

of the PGS is the interval, measured in the standard linear

manner, between two elements or points as they would exist

in a two-dimensional plane (Gutman, 1966, p. 106; Michelson,

1970, p. 48). Physical distance is a less realistic meas-

ure, in an urban context, than functional distance (Gustmin,

1966, p. 106; Michelson, 1970, p. 173), the distance actu-

ally traversed. As a modified physical distance measure,

functional distance is, however, somewhat myopic in thpt it

neglects the potentially more relevant time (or possible

monetary costs) involved in spatial movements. For this

reason it has been suggested that time is a useful and

realistic measure of functional distance (Gist and Halbert,

1956, p. 78; Isard, 1960, p. 206; B. Thompson, 1967; Yuill,

1967, p. 113; Brunner and Mason, 1968; Cox and Cooke, 1970),

in a fashion similar to that involved in measuring economic

distance (Deutsch and Isard, 1961).

This latter version is not a totally adequate measure

of the absolute distance component of the MGS (mental

geographic set) for four reasons. First, as discussed, the

costs associated with customer movements are not confined

to merely time or money and are not what might be considered

transport costs. Second, certain potentially relevant costs

may not even enter into a given customer's environmental

decision process. Third, the costs which are relevant will

tend to vary with the environmental situation and depend, for

instance, on available transport mode or a customer's

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs